url
stringlengths
32
34
url_pdf
stringlengths
32
34
title
stringlengths
9
197
authors
stringlengths
6
601
primary_category
stringclasses
1 value
categories
stringlengths
5
61
abstract
stringlengths
34
3.18k
updated
timestamp[s]
published
timestamp[s]
embeddings_title
list
embeddings_abstract
list
http://arxiv.org/abs/2006.00490v1
http://arxiv.org/pdf/2006.00490v1
Detecting Group Beliefs Related to 2018's Brazilian Elections in Tweets A Combined Study on Modeling Topics and Sentiment Analysis
Brenda Salenave Santana, Aline Aver Vanin
cs.CL
cs.CL
2018's Brazilian presidential elections highlighted the influence of alternative media and social networks, such as Twitter. In this work, we perform an analysis covering politically motivated discourses related to the second round in Brazilian elections. In order to verify whether similar discourses reinforce group en...
2020-05-31T10:58:35
2020-05-31T10:58:35
[ 0.036499023, -0.0012273788, 0.004951477, -0.03225708, -0.008255005, 0.019256592, -0.027801514, -0.0060653687, 0.016220093, 0.037353516, -0.020523071, -0.0037555695, 0.018066406, -0.013061523, -0.030090332, -0.017456055, -0.011795044, 0.050628662, 0.028152466, -0.02142334, -0....
[ 0.042297363, -0.025222778, 0.008163452, -0.0143966675, -0.0023384094, -0.0039100647, -0.022369385, -0.018539429, -0.009559631, 0.04324341, -0.011428833, 0.006084442, 0.028701782, 0.00032424927, -0.0036830902, -0.03152466, -0.035461426, 0.0637207, 0.045898438, -0.02128601, -0....
http://arxiv.org/abs/2006.00512v1
http://arxiv.org/pdf/2006.00512v1
Learning to Recognise Words using Visually Grounded Speech
Sebastiaan Scholten, Danny Merkx, Odette Scharenborg
cs.CL
cs.CL
We investigated word recognition in a Visually Grounded Speech model. The model has been trained on pairs of images and spoken captions to create visually grounded embeddings which can be used for speech to image retrieval and vice versa. We investigate whether such a model can be used to recognise words by embedding i...
2020-05-31T12:48:37
2020-05-31T12:48:37
[ 0.0132751465, 0.014831543, -0.037719727, -0.042266846, -0.024658203, -0.008811951, -0.03765869, 0.0033035278, 0.042419434, 0.03793335, -0.05001831, -0.000040709972, 0.0052948, -0.0018987656, -0.06677246, -0.03161621, -0.0033054352, 0.07757568, 0.020751953, -0.00094890594, -0....
[ 0.049224854, 0.022827148, -0.0031929016, 0.005088806, -0.021972656, -0.040252686, -0.011405945, -0.032287598, 0.01096344, 0.058776855, -0.044555664, 0.012588501, 0.0357666, 0.0132751465, -0.030014038, -0.016281128, -0.0073509216, 0.07824707, 0.028640747, -0.013374329, -0.0170...
http://arxiv.org/abs/2006.00575v4
http://arxiv.org/pdf/2006.00575v4
Neural Entity Linking: A Survey of Models Based on Deep Learning
Ozge Sevgili, Artem Shelmanov, Mikhail Arkhipov, Alexander Panchenko, Chris Biemann
cs.CL
cs.CL
This survey presents a comprehensive description of recent neural entity linking (EL) systems developed since 2015 as a result of the "deep learning revolution" in natural language processing. Its goal is to systemize design features of neural entity linking systems and compare their performance to the remarkable class...
2022-04-07T17:56:30
2020-05-31T18:02:26
[ 0.026550293, 0.011962891, -0.03668213, -0.0446167, 0.0021591187, -0.016983032, -0.023147583, -0.013679504, 0.020980835, 0.05279541, 0.026855469, -0.0075531006, 0.06021118, -0.043151855, 0.0071525574, 0.024993896, -0.024047852, 0.03201294, -0.014823914, -0.01576233, 0.01446533...
[ 0.036193848, -0.0013437271, -0.020462036, -0.0099105835, 0.012527466, -0.013191223, -0.02607727, -0.012825012, 0.006462097, 0.04788208, -0.0004029274, 0.0052757263, 0.08215332, -0.011444092, 0.03768921, 0.005645752, -0.023910522, 0.052947998, 0.0034217834, -0.021499634, -0.01...
http://arxiv.org/abs/2006.00578v1
http://arxiv.org/pdf/2006.00578v1
"Judge me by my size (noun), do you?'' YodaLib: A Demographic-Aware Humor Generation Framework
Aparna Garimella, Carmen Banea, Nabil Hossain, Rada Mihalcea
cs.CL
cs.CL
The subjective nature of humor makes computerized humor generation a challenging task. We propose an automatic humor generation framework for filling the blanks in Mad Libs stories, while accounting for the demographic backgrounds of the desired audience. We collect a dataset consisting of such stories, which are fille...
2020-05-31T18:11:52
2020-05-31T18:11:52
[ 0.02192688, -0.026367188, -0.030029297, -0.029724121, -0.02859497, -0.004234314, -0.024276733, 0.008644104, 0.018600464, 0.029571533, -0.01600647, -0.004699707, -0.02609253, -0.019866943, -0.05050659, -0.020736694, -0.004634857, -0.007850647, -0.03289795, -0.041381836, 0.0087...
[ 0.052856445, -0.021530151, 0.015670776, -0.014274597, -0.022323608, -0.016799927, -0.022277832, -0.033721924, 0.038879395, 0.033691406, 0.0010299683, -0.009155273, 0.023712158, -0.0033607483, -0.028015137, 0.001326561, -0.028030396, 0.034118652, -0.0075950623, -0.05810547, 0....
http://arxiv.org/abs/2006.00591v2
http://arxiv.org/pdf/2006.00591v2
Efficient Deployment of Conversational Natural Language Interfaces over Databases
Anthony Colas, Trung Bui, Franck Dernoncourt, Moumita Sinha, Doo Soon Kim
cs.CL
cs.CL
Many users communicate with chatbots and AI assistants in order to help them with various tasks. A key component of the assistant is the ability to understand and answer a user's natural language questions for question-answering (QA). Because data can be usually stored in a structured manner, an essential step involves...
2020-06-04T19:31:14
2020-05-31T19:16:27
[ 0.01159668, -0.05026245, -0.042907715, -0.082336426, -0.017471313, 0.0013523102, -0.055419922, 0.04827881, 0.06506348, 0.02243042, -0.040374756, -0.034820557, 0.0004146099, -0.022827148, -0.060760498, -0.03817749, -0.03277588, 0.058685303, -0.0184021, -0.037597656, 0.00884246...
[ 0.022644043, -0.045776367, -0.0069732666, -0.008605957, -0.026443481, -0.0059814453, -0.04034424, -0.0034923553, 0.055236816, 0.032684326, -0.027267456, -0.007858276, 0.014762878, 0.004676819, -0.039154053, -0.031021118, -0.014297485, 0.04751587, -0.015533447, -0.034576416, -...
http://arxiv.org/abs/2006.00593v2
http://arxiv.org/pdf/2006.00593v2
BPGC at SemEval-2020 Task 11: Propaganda Detection in News Articles with Multi-Granularity Knowledge Sharing and Linguistic Features based Ensemble Learning
Rajaswa Patil, Somesh Singh, Swati Agarwal
cs.CL
cs.CL
Propaganda spreads the ideology and beliefs of like-minded people, brainwashing their audiences, and sometimes leading to violence. SemEval 2020 Task-11 aims to design automated systems for news propaganda detection. Task-11 consists of two sub-tasks, namely, Span Identification - given any news article, the system tag...
2020-08-24T12:34:38
2020-05-31T19:35:53
[ 0.038879395, 0.00390625, -0.002922058, -0.010719299, 0.0006709099, -0.006839752, -0.028442383, -0.0625, -0.013771057, 0.017471313, -0.012336731, 0.01499176, -0.0077590942, -0.01259613, -0.0007901192, -0.037261963, -0.033966064, 0.06939697, -0.0037288666, -0.02796936, -0.03894...
[ 0.068847656, 0.012886047, 0.014541626, 0.005592346, 0.019241333, -0.024734497, 0.007160187, -0.0703125, -0.014846802, 0.019973755, -0.019180298, 0.02609253, 0.0072669983, -0.003232956, 0.022323608, -0.042541504, -0.031555176, 0.097229004, 0.03878784, -0.025268555, -0.0234375,...
http://arxiv.org/abs/2006.00607v1
http://arxiv.org/pdf/2006.00607v1
LRG at SemEval-2020 Task 7: Assessing the Ability of BERT and Derivative Models to Perform Short-Edits based Humor Grading
Siddhant Mahurkar, Rajaswa Patil
cs.CL
cs.CL
In this paper, we assess the ability of BERT and its derivative models (RoBERTa, DistilBERT, and ALBERT) for short-edits based humor grading. We test these models for humor grading and classification tasks on the Humicroedit and the FunLines dataset. We perform extensive experiments with these models to test their lang...
2020-05-31T20:55:08
2020-05-31T20:55:08
[ 0.03640747, -0.018615723, 0.0006351471, -0.023971558, -0.02330017, -0.0064582825, -0.028518677, -0.011489868, 0.04473877, 0.033477783, 0.001537323, -0.029388428, 0.024520874, -0.01939392, -0.041748047, 0.005493164, -0.04748535, 0.036254883, -0.054473877, -0.04574585, 0.008468...
[ 0.042419434, -0.018051147, 0.00932312, -0.018844604, -0.03100586, -0.017410278, -0.037353516, -0.01550293, 0.04156494, 0.052246094, 0.0017881393, -0.035125732, 0.047058105, -0.000012040138, -0.05316162, -0.0043563843, -0.045898438, 0.07763672, -0.008079529, -0.031066895, 0.01...
http://arxiv.org/abs/2006.00609v1
http://arxiv.org/pdf/2006.00609v1
CNRL at SemEval-2020 Task 5: Modelling Causal Reasoning in Language with Multi-Head Self-Attention Weights based Counterfactual Detection
Rajaswa Patil, Veeky Baths
cs.CL
cs.CL
In this paper, we describe an approach for modelling causal reasoning in natural language by detecting counterfactuals in text using multi-head self-attention weights. We use pre-trained transformer models to extract contextual embeddings and self-attention weights from the text. We show the use of convolutional layers...
2020-05-31T21:02:25
2020-05-31T21:02:25
[ 0.038482666, -0.023757935, -0.018173218, -0.058746338, -0.038269043, 0.00894165, -0.072387695, 0.0045547485, 0.04006958, 0.00058174133, -0.008796692, -0.012084961, 0.03286743, -0.039154053, 0.008262634, -0.0052948, -0.011604309, 0.06274414, 0.022537231, -0.020828247, -0.00575...
[ 0.038513184, -0.0073432922, -0.021270752, -0.04852295, -0.019592285, -0.0020942688, -0.06149292, -0.005203247, 0.02619934, 0.0112838745, -0.024017334, -0.0069274902, 0.06317139, -0.009674072, 0.015335083, -0.022338867, -0.010154724, 0.08282471, 0.04321289, -0.018508911, -0.01...
http://arxiv.org/abs/2006.00632v2
http://arxiv.org/pdf/2006.00632v2
Neural Unsupervised Domain Adaptation in NLP---A Survey
Alan Ramponi, Barbara Plank
cs.CL
cs.CL
Deep neural networks excel at learning from labeled data and achieve state-of-the-art resultson a wide array of Natural Language Processing tasks. In contrast, learning from unlabeled data, especially under domain shift, remains a challenge. Motivated by the latest advances, in this survey we review neural unsupervised...
2020-10-28T08:24:14
2020-05-31T22:34:14
[ -0.0066947937, 0.0066070557, -0.031982422, -0.0075149536, -0.021896362, -0.014228821, -0.027877808, -0.0050315857, 0.027832031, 0.026397705, 0.0050086975, -0.011520386, 0.017196655, -0.015792847, -0.0073661804, -0.001912117, -0.029418945, 0.05822754, -0.02911377, -0.037719727, ...
[ 0.027816772, -0.0057411194, -0.032226562, 0.0028953552, -0.011024475, -0.026473999, -0.036346436, -0.021865845, 0.031951904, 0.025726318, -0.0056877136, 0.00094747543, 0.010101318, -0.019546509, 0.025253296, -0.018493652, -0.012672424, 0.059753418, -0.008834839, -0.034729004, ...
http://arxiv.org/abs/2006.00635v2
http://arxiv.org/pdf/2006.00635v2
A Unified Feature Representation for Lexical Connotations
Emily Allaway, Kathleen McKeown
cs.CL
cs.CL
Ideological attitudes and stance are often expressed through subtle meanings of words and phrases. Understanding these connotations is critical to recognizing the cultural and emotional perspectives of the speaker. In this paper, we use distant labeling to create a new lexical resource representing connotation aspects ...
2021-03-01T14:14:21
2020-05-31T23:14:02
[ 0.034729004, -0.027359009, 0.009063721, -0.026321411, -0.042785645, 0.0011234283, -0.053649902, -0.015464783, 0.042022705, 0.049926758, 0.0011692047, -0.035308838, 0.048095703, -0.010002136, -0.028915405, 0.0032634735, -0.010559082, 0.051971436, 0.021530151, -0.033355713, -0....
[ 0.05267334, -0.017318726, 0.03503418, -0.003873825, -0.020248413, -0.02079773, -0.01953125, -0.038604736, 0.032348633, 0.011932373, -0.050445557, -0.012290955, 0.02128601, 0.0037994385, -0.0034236908, -0.041137695, -0.00699234, 0.0491333, 0.03842163, -0.020874023, -0.04010009...
http://arxiv.org/abs/2006.00671v2
http://arxiv.org/pdf/2006.00671v2
Conversational Machine Comprehension: a Literature Review
Somil Gupta, Bhanu Pratap Singh Rawat, Hong Yu
cs.CL
cs.CL
Conversational Machine Comprehension (CMC), a research track in conversational AI, expects the machine to understand an open-domain natural language text and thereafter engage in a multi-turn conversation to answer questions related to the text. While most of the research in Machine Reading Comprehension (MRC) revolves...
2020-11-05T06:28:06
2020-06-01T02:20:08
[ -0.029159546, -0.030776978, 0.007522583, -0.018051147, -0.07269287, 0.023864746, -0.08105469, -0.016479492, 0.04168701, 0.054901123, -0.03805542, -0.016555786, 0.043823242, 0.0023651123, -0.05834961, 0.005191803, -0.035003662, 0.037719727, -0.03781128, -0.02760315, -0.0080795...
[ -0.0019760132, -0.023162842, 0.009796143, -0.0058670044, -0.042755127, 0.010612488, -0.0680542, -0.033477783, 0.051696777, 0.032104492, -0.053588867, -0.00737381, 0.033050537, 0.009963989, -0.039276123, -0.02128601, -0.029373169, 0.04031372, -0.022064209, -0.03704834, -0.0164...
http://arxiv.org/abs/2006.00814v3
http://arxiv.org/pdf/2006.00814v3
Online Versus Offline NMT Quality: An In-depth Analysis on English-German and German-English
Maha Elbayad, Michael Ustaszewski, Emmanuelle Esperança-Rodier, Francis Brunet Manquat, Jakob Verbeek, Laurent Besacier
cs.CL
cs.CL
We conduct in this work an evaluation study comparing offline and online neural machine translation architectures. Two sequence-to-sequence models: convolutional Pervasive Attention (Elbayad et al. 2018) and attention-based Transformer (Vaswani et al. 2017) are considered. We investigate, for both architectures, the im...
2020-11-24T09:10:30
2020-06-01T09:43:54
[ 0.022232056, -0.023010254, 0.0395813, -0.013542175, -0.03479004, 0.029006958, -0.008628845, -0.021087646, -0.013244629, 0.052246094, 0.04324341, 0.013450623, 0.041748047, -0.043029785, 0.016174316, 0.017044067, -0.032043457, 0.027359009, -0.0035057068, -0.02897644, 0.01757812...
[ 0.016403198, -0.025177002, 0.036376953, 0.016708374, -0.057647705, 0.015777588, -0.02407837, -0.03439331, 0.027008057, 0.049316406, -0.0038700104, -0.03277588, 0.07635498, -0.013298035, -0.021209717, 0.0025348663, -0.0014762878, 0.047088623, 0.013000488, -0.03277588, -0.01308...
http://arxiv.org/abs/2006.00838v1
http://arxiv.org/pdf/2006.00838v1
Efficient EUD Parsing
Mathieu Dehouck, Mark Anderson, Carlos Gómez-Rodríguez
cs.CL
cs.CL
We present the system submission from the FASTPARSE team for the EUD Shared Task at IWPT 2020. We engaged with the task by focusing on efficiency. For this we considered training costs and inference efficiency. Our models are a combination of distilled neural dependency parsers and a rule-based system that projects UD ...
2020-06-01T10:31:56
2020-06-01T10:31:56
[ 0.010612488, 0.009056091, 0.0014123917, -0.047576904, 0.01776123, 0.015167236, -0.016403198, -0.024887085, 0.020111084, 0.045318604, 0.031585693, 0.010826111, 0.011734009, -0.03439331, -0.02633667, 0.0236969, -0.03326416, 0.054901123, -0.0015029907, -0.031051636, 0.031707764,...
[ 0.03074646, -0.006385803, -0.0032138824, -0.04498291, -0.018615723, 0.012207031, -0.055786133, -0.008804321, 0.020965576, 0.029937744, -0.014053345, -0.013069153, 0.010231018, -0.03869629, -0.0025634766, -0.0074501038, -0.0063056946, 0.054901123, 0.025131226, -0.030715942, 0....
http://arxiv.org/abs/2006.00843v2
http://arxiv.org/pdf/2006.00843v2
Rhetoric, Logic, and Dialectic: Advancing Theory-based Argument Quality Assessment in Natural Language Processing
Anne Lauscher, Lily Ng, Courtney Napoles, Joel Tetreault
cs.CL
cs.CL
Though preceding work in computational argument quality (AQ) mostly focuses on assessing overall AQ, researchers agree that writers would benefit from feedback targeting individual dimensions of argumentation theory. However, a large-scale theory-based corpus and corresponding computational models are missing. We fill ...
2020-11-03T09:26:07
2020-06-01T10:39:50
[ 0.04837036, -0.019088745, -0.014549255, -0.05432129, -0.015258789, 0.012931824, -0.03704834, -0.000037133694, -0.033050537, 0.043823242, 0.0053482056, 0.027160645, -0.002368927, -0.0017976761, -0.009963989, -0.009796143, -0.032470703, 0.05645752, -0.022872925, -0.05029297, -0...
[ 0.050842285, -0.019348145, 0.019073486, -0.037506104, -0.0074005127, 0.015289307, -0.01727295, -0.027359009, 0.01020813, 0.021377563, 0.027664185, 0.013786316, 0.011436462, -0.019577026, 0.0026569366, -0.018600464, -0.029296875, 0.06854248, -0.009841919, -0.02758789, -0.00897...
http://arxiv.org/abs/2006.00844v1
http://arxiv.org/pdf/2006.00844v1
Distilling Neural Networks for Greener and Faster Dependency Parsing
Mark Anderson, Carlos Gómez-Rodríguez
cs.CL
cs.CL
The carbon footprint of natural language processing research has been increasing in recent years due to its reliance on large and inefficient neural network implementations. Distillation is a network compression technique which attempts to impart knowledge from a large model to a smaller one. We use teacher-student dis...
2020-06-01T10:43:53
2020-06-01T10:43:53
[ 0.009864807, -0.008964539, -0.0637207, -0.10015869, -0.008293152, 0.017044067, -0.09454346, 0.0440979, 0.019882202, 0.018936157, -0.026260376, -0.013618469, -0.047027588, -0.011299133, -0.027175903, -0.040618896, -0.0040893555, 0.03564453, -0.0027389526, -0.031402588, -0.0005...
[ 0.00491333, -0.018661499, -0.033416748, -0.06378174, -0.02720642, 0.013046265, -0.09448242, 0.014778137, 0.03390503, 0.00944519, -0.03744507, -0.018997192, -0.008071899, -0.008773804, -0.019165039, -0.04360962, -0.017547607, 0.036224365, -0.00642395, -0.035614014, 0.022918701...
http://arxiv.org/abs/2006.00995v3
http://arxiv.org/pdf/2006.00995v3
Amnesic Probing: Behavioral Explanation with Amnesic Counterfactuals
Yanai Elazar, Shauli Ravfogel, Alon Jacovi, Yoav Goldberg
cs.CL
cs.CL
A growing body of work makes use of probing to investigate the working of neural models, often considered black boxes. Recently, an ongoing debate emerged surrounding the limitations of the probing paradigm. In this work, we point out the inability to infer behavioral conclusions from probing results and offer an alter...
2021-02-19T09:01:59
2020-06-01T15:00:11
[ 0.037078857, -0.012077332, 0.05633545, 0.0211792, -0.051635742, 0.00008249283, 0.02684021, -0.049316406, -0.007247925, 0.016815186, -0.008041382, -0.008857727, -0.016204834, -0.03543091, 0.0010375977, -0.028579712, -0.0027065277, 0.016281128, 0.0016355515, -0.021484375, -0.01...
[ 0.012565613, -0.02583313, 0.02848816, 0.020477295, -0.055847168, -0.018310547, 0.008651733, -0.029403687, -0.009712219, 0.03955078, 0.012916565, -0.0023784637, 0.039886475, -0.02456665, 0.00491333, -0.018508911, -0.018798828, 0.049987793, 0.021194458, -0.022994995, -0.0160217...
http://arxiv.org/abs/2006.00998v1
http://arxiv.org/pdf/2006.00998v1
Toxicity Detection: Does Context Really Matter?
John Pavlopoulos, Jeffrey Sorensen, Lucas Dixon, Nithum Thain, Ion Androutsopoulos
cs.CL
cs.CL
Moderation is crucial to promoting healthy on-line discussions. Although several `toxicity' detection datasets and models have been published, most of them ignore the context of the posts, implicitly assuming that comments maybe judged independently. We investigate this assumption by focusing on two questions: (a) does...
2020-06-01T15:03:48
2020-06-01T15:03:48
[ 0.053588867, 0.015045166, 0.014541626, -0.028640747, 0.0028133392, -0.006046295, -0.0006637573, -0.031982422, -0.015899658, 0.024047852, -0.035095215, 0.012710571, -0.02067566, -0.010902405, 0.005821228, -0.048034668, -0.022613525, 0.058807373, 0.028884888, -0.016174316, -0.0...
[ 0.09338379, 0.0052833557, 0.018798828, -0.028930664, 0.030075073, -0.02217102, 0.027313232, -0.041107178, 0.008583069, 0.017715454, -0.045196533, 0.035827637, 0.0011949539, 0.011138916, 0.008560181, -0.039520264, -0.00944519, 0.054504395, 0.038513184, -0.024765015, -0.0138244...
http://arxiv.org/abs/2006.01038v3
http://arxiv.org/pdf/2006.01038v3
DocBank: A Benchmark Dataset for Document Layout Analysis
Minghao Li, Yiheng Xu, Lei Cui, Shaohan Huang, Furu Wei, Zhoujun Li, Ming Zhou
cs.CL
cs.CL
Document layout analysis usually relies on computer vision models to understand documents while ignoring textual information that is vital to capture. Meanwhile, high quality labeled datasets with both visual and textual information are still insufficient. In this paper, we present \textbf{DocBank}, a benchmark dataset...
2020-11-11T05:08:05
2020-06-01T16:04:30
[ -0.051727295, -0.039215088, -0.050079346, -0.051818848, -0.06677246, 0.01675415, -0.114868164, 0.0067367554, 0.051635742, 0.057434082, -0.036956787, 0.013969421, 0.020706177, 0.0001924038, -0.03817749, -0.0015258789, 0.022537231, -0.0042800903, 0.0025558472, -0.009597778, -0....
[ -0.026519775, -0.032043457, -0.022735596, -0.014076233, -0.06793213, 0.002752304, -0.10626221, -0.015777588, 0.06652832, 0.03918457, -0.028366089, 0.0074157715, 0.037200928, 0.005088806, -0.0036315918, -0.0048332214, 0.02331543, -0.004081726, -0.0053901672, -0.010848999, -0.0...
http://arxiv.org/abs/2006.01080v1
http://arxiv.org/pdf/2006.01080v1
Is 42 the Answer to Everything in Subtitling-oriented Speech Translation?
Alina Karakanta, Matteo Negri, Marco Turchi
cs.CL
cs.CL
Subtitling is becoming increasingly important for disseminating information, given the enormous amounts of audiovisual content becoming available daily. Although Neural Machine Translation (NMT) can speed up the process of translating audiovisual content, large manual effort is still required for transcribing the sourc...
2020-06-01T17:02:28
2020-06-01T17:02:28
[ -0.01159668, -0.031463623, 0.02670288, -0.04067993, -0.0368042, 0.020187378, -0.025482178, 0.03137207, 0.025253296, 0.036102295, 0.0022087097, -0.036987305, 0.018920898, -0.05203247, -0.034729004, -0.020187378, -0.0016050339, 0.002374649, -0.05697632, -0.015472412, -0.0250091...
[ 0.004261017, -0.0073928833, 0.01184845, 0.0019931793, -0.010269165, 0.009597778, -0.041015625, -0.030136108, -0.002483368, 0.06500244, 0.01108551, -0.020935059, 0.048919678, -0.02180481, -0.01398468, -0.02381897, -0.0038013458, 0.01260376, -0.0059661865, -0.032836914, -0.0033...
http://arxiv.org/abs/2006.01117v1
http://arxiv.org/pdf/2006.01117v1
NSTM: Real-Time Query-Driven News Overview Composition at Bloomberg
Joshua Bambrick, Minjie Xu, Andy Almonte, Igor Malioutov, Guim Perarnau, Vittorio Selo, Iat Chong Chan
cs.CL
cs.CL
Millions of news articles from hundreds of thousands of sources around the globe appear in news aggregators every day. Consuming such a volume of news presents an almost insurmountable challenge. For example, a reader searching on Bloomberg's system for news about the U.K. would find 10,000 articles on a typical day. A...
2020-06-01T17:58:57
2020-06-01T17:58:57
[ 0.0043945312, -0.033447266, -0.041290283, -0.10003662, -0.01576233, 0.011245728, -0.049926758, 0.03793335, 0.03451538, 0.05834961, -0.020950317, -0.003704071, -0.0061416626, -0.004184723, -0.06530762, -0.014602661, 0.0018415451, -0.005378723, -0.021224976, -0.028701782, 0.007...
[ 0.023925781, -0.018035889, -0.021942139, -0.058685303, -0.021774292, 0.00009483099, -0.064453125, -0.003982544, 0.0121154785, 0.0574646, 0.015274048, 0.012237549, -0.011306763, 0.0034866333, -0.035888672, 0.0017242432, 0.011154175, -0.0039024353, 0.0078125, -0.030044556, 0.01...
http://arxiv.org/abs/2006.01175v2
http://arxiv.org/pdf/2006.01175v2
Lexical Normalization for Code-switched Data and its Effect on POS-tagging
Rob van der Goot, Özlem Çetinoğlu
cs.CL
cs.CL
Lexical normalization, the translation of non-canonical data to standard language, has shown to improve the performance of manynatural language processing tasks on social media. Yet, using multiple languages in one utterance, also called code-switching (CS), is frequently overlooked by these normalization systems, desp...
2021-01-31T20:57:52
2020-06-01T18:07:15
[ 0.006477356, -0.011459351, 0.0042037964, 0.005493164, -0.016143799, -0.0010957718, 0.0013828278, -0.008392334, 0.012969971, 0.032592773, -0.046722412, 0.016708374, 0.024856567, 0.0018644333, 0.0026130676, 0.002538681, -0.009132385, 0.023483276, -0.015792847, -0.020874023, 0.0...
[ 0.05026245, -0.01687622, 0.014076233, -0.0020122528, -0.0075302124, -0.010032654, -0.02508545, -0.023956299, -0.0030937195, 0.036895752, -0.010986328, 0.030181885, 0.024047852, -0.00023162365, 0.008934021, -0.012382507, -0.020111084, 0.084106445, 0.026473999, -0.01600647, -0....
http://arxiv.org/abs/2006.01189v1
http://arxiv.org/pdf/2006.01189v1
An Effective Contextual Language Modeling Framework for Speech Summarization with Augmented Features
Shi-Yan Weng, Tien-Hong Lo, Berlin Chen
cs.CL
cs.CL
Tremendous amounts of multimedia associated with speech information are driving an urgent need to develop efficient and effective automatic summarization methods. To this end, we have seen rapid progress in applying supervised deep neural network-based methods to extractive speech summarization. More recently, the Bidi...
2020-06-01T18:27:48
2020-06-01T18:27:48
[ 0.044647217, 0.005306244, -0.023742676, -0.089416504, -0.018051147, 0.026428223, -0.12145996, -0.0206604, 0.007789612, 0.08868408, 0.032226562, -0.009361267, 0.053619385, -0.011001587, -0.045684814, 0.014480591, 0.0013856888, 0.025100708, -0.022918701, -0.026870728, -0.011581...
[ 0.051116943, -0.0043144226, 0.005657196, -0.06921387, -0.014266968, 0.0075950623, -0.08154297, -0.0119018555, 0.025222778, 0.056854248, -0.018234253, 0.024032593, 0.040649414, 0.003944397, -0.011932373, -0.0039901733, -0.009124756, 0.04824829, 0.0026512146, -0.017333984, -0.0...
http://arxiv.org/abs/2006.01205v2
http://arxiv.org/pdf/2006.01205v2
CS-NLP team at SemEval-2020 Task 4: Evaluation of State-of-the-art NLP Deep Learning Architectures on Commonsense Reasoning Task
Sirwe Saeedi, Aliakbar Panahi, Seyran Saeedi, Alvis C Fong
cs.CL
cs.CL
In this paper, we investigate a commonsense inference task that unifies natural language understanding and commonsense reasoning. We describe our attempt at SemEval-2020 Task 4 competition: Commonsense Validation and Explanation (ComVE) challenge. We discuss several state-of-the-art deep learning architectures for this...
2020-07-18T15:19:20
2020-05-17T13:20:10
[ 0.023910522, -0.04537964, -0.0049476624, -0.023345947, -0.04562378, 0.0064926147, -0.022109985, 0.008491516, 0.06121826, 0.017623901, -0.023117065, -0.009185791, 0.017105103, -0.022109985, -0.0035095215, 0.0018224716, -0.019485474, 0.058807373, -0.009635925, -0.028381348, 0.0...
[ 0.03048706, -0.035491943, 0.006160736, -0.025604248, -0.035125732, 0.0146865845, -0.05419922, -0.011756897, 0.03640747, 0.0030345917, -0.0057678223, 0.00042653084, 0.004699707, -0.0077705383, 0.0041275024, -0.024841309, -0.022369385, 0.0619812, 0.011505127, -0.031173706, 0.01...
http://arxiv.org/abs/2006.01210v1
http://arxiv.org/pdf/2006.01210v1
The 'Letter' Distribution in the Chinese Language
Qinghua Chen, Yan Wang, Mengmeng Wang, Xiaomeng Li
cs.CL
cs.CL, 62P25, 97M80
Corpus-based statistical analysis plays a significant role in linguistic research, and ample evidence has shown that different languages exhibit some common laws. Studies have found that letters in some alphabetic writing languages have strikingly similar statistical usage frequency distributions. Does this hold for Ch...
2020-05-26T05:18:56
2020-05-26T05:18:56
[ -0.066833496, -0.0015716553, -0.004184723, 0.060791016, -0.030075073, -0.0060424805, 0.015808105, -0.018310547, 0.035339355, 0.045135498, -0.005077362, -0.0023155212, 0.014305115, -0.03173828, -0.027404785, 0.02293396, -0.017211914, 0.03918457, 0.00043153763, -0.01612854, 0.0...
[ -0.0031852722, -0.008743286, -0.005695343, 0.037139893, -0.0059928894, 0.001244545, -0.0093307495, -0.024887085, 0.041015625, 0.04736328, -0.023986816, 0.006465912, 0.02861023, -0.005970001, -0.019226074, 0.0032901764, -0.025634766, 0.07965088, 0.02986145, -0.029785156, 0.011...
http://arxiv.org/abs/2006.01266v1
http://arxiv.org/pdf/2006.01266v1
Leveraging Affective Bidirectional Transformers for Offensive Language Detection
AbdelRahim Elmadany, Chiyu Zhang, Muhammad Abdul-Mageed, Azadeh Hashemi
cs.CL
cs.CL
Social media are pervasive in our life, making it necessary to ensure safe online experiences by detecting and removing offensive and hate speech. In this work, we report our submission to the Offensive Language and hate-speech Detection shared task organized with the 4th Workshop on Open-Source Arabic Corpora and Proc...
2020-05-16T04:55:35
2020-05-16T04:55:35
[ 0.05895996, 0.012641907, 0.005886078, -0.034301758, -0.011039734, -0.008811951, -0.008056641, -0.038879395, 0.062561035, 0.026397705, -0.05203247, -0.02104187, 0.025527954, 0.00047826767, -0.03564453, -0.038513184, -0.014007568, 0.070251465, 0.023712158, -0.002275467, -0.0119...
[ 0.06427002, 0.000058948994, 0.012939453, -0.035186768, 0.006099701, -0.029968262, -0.014350891, -0.024459839, 0.03894043, 0.011108398, -0.010879517, -0.014465332, 0.024108887, -0.016616821, 0.0052108765, -0.008514404, -0.005832672, 0.056915283, 0.028167725, -0.021118164, -0.0...
http://arxiv.org/abs/2006.01338v1
http://arxiv.org/pdf/2006.01338v1
A Survey of Neural Networks and Formal Languages
Joshua Ackerman, George Cybenko
cs.CL
cs.CL
This report is a survey of the relationships between various state-of-the-art neural network architectures and formal languages as, for example, structured by the Chomsky Language Hierarchy. Of particular interest are the abilities of a neural architecture to represent, recognize and generate words from a specific lang...
2020-06-02T01:46:04
2020-06-02T01:46:04
[ -0.011116028, -0.02293396, -0.018051147, -0.014755249, -0.035369873, -0.008308411, -0.043914795, 0.013244629, 0.0054397583, 0.02558899, -0.0001834631, -0.004814148, -0.015014648, -0.012374878, -0.02355957, -0.012390137, -0.012145996, 0.04220581, -0.007133484, -0.024902344, -0...
[ 0.03390503, -0.016708374, 0.012458801, -0.010116577, -0.03552246, 0.0040016174, -0.041809082, -0.017501831, 0.010223389, 0.04852295, -0.015686035, -0.021514893, 0.021453857, -0.0036315918, -0.03112793, -0.00844574, 0.0027313232, 0.040740967, 0.036071777, -0.026123047, 0.00648...
http://arxiv.org/abs/2006.01372v2
http://arxiv.org/pdf/2006.01372v2
Embeddings of Label Components for Sequence Labeling: A Case Study of Fine-grained Named Entity Recognition
Takuma Kato, Kaori Abe, Hiroki Ouchi, Shumpei Miyawaki, Jun Suzuki, Kentaro Inui
cs.CL
cs.CL
In general, the labels used in sequence labeling consist of different types of elements. For example, IOB-format entity labels, such as B-Person and I-Person, can be decomposed into span (B and I) and type information (Person). However, while most sequence labeling models do not consider such label components, the shar...
2020-06-04T14:01:18
2020-06-02T03:47:19
[ 0.018722534, -0.01576233, -0.028182983, -0.050842285, -0.0057792664, 0.007144928, -0.03842163, 0.030395508, 0.027908325, 0.020492554, -0.04135132, -0.008880615, 0.014373779, -0.011016846, -0.020751953, -0.029708862, -0.0064353943, 0.07159424, 0.0013046265, -0.03050232, -0.053...
[ 0.03967285, -0.01486969, -0.0146484375, -0.05218506, -0.016815186, -0.0048065186, -0.06414795, 0.0077285767, 0.03994751, 0.0129852295, -0.05340576, 0.013839722, 0.04284668, 0.002161026, 0.0065689087, -0.014831543, -0.015281677, 0.07299805, 0.021820068, -0.021652222, -0.025451...
http://arxiv.org/abs/2006.01432v1
http://arxiv.org/pdf/2006.01432v1
BERT Based Multilingual Machine Comprehension in English and Hindi
Somil Gupta, Nilesh Khade
cs.CL
cs.CL
Multilingual Machine Comprehension (MMC) is a Question-Answering (QA) sub-task that involves quoting the answer for a question from a given snippet, where the question and the snippet can be in different languages. Recently released multilingual variant of BERT (m-BERT), pre-trained with 104 languages, has performed we...
2020-06-02T07:36:49
2020-06-02T07:36:49
[ -0.03994751, -0.05218506, 0.017105103, -0.06518555, -0.060424805, 0.054626465, -0.10369873, 0.032562256, 0.03173828, 0.05441284, -0.030563354, -0.007965088, 0.021118164, -0.004776001, -0.03540039, 0.00548172, -0.039611816, 0.038146973, -0.0184021, -0.015838623, -0.0020542145,...
[ -0.0067214966, -0.057495117, 0.0057868958, -0.018981934, -0.053466797, 0.010765076, -0.07885742, 0.0077934265, 0.031280518, 0.036468506, -0.018539429, -0.016235352, 0.033294678, 0.0006980896, -0.023757935, 0.0016870499, -0.015899658, 0.056762695, 0.015487671, -0.019805908, -0...
http://arxiv.org/abs/2006.01554v2
http://arxiv.org/pdf/2006.01554v2
A Contextual Hierarchical Attention Network with Adaptive Objective for Dialogue State Tracking
Yong Shan, Zekang Li, Jinchao Zhang, Fandong Meng, Yang Feng, Cheng Niu, Jie Zhou
cs.CL
cs.CL
Recent studies in dialogue state tracking (DST) leverage historical information to determine states which are generally represented as slot-value pairs. However, most of them have limitations to efficiently exploit relevant context due to the lack of a powerful mechanism for modeling interactions between the slot and t...
2020-12-21T02:03:40
2020-06-02T12:25:44
[ 0.037139893, 0.00017547607, -0.034851074, -0.077819824, -0.026489258, -0.0053977966, -0.06185913, 0.004425049, 0.027511597, 0.029464722, 0.009666443, -0.024353027, 0.026641846, -0.028030396, -0.007896423, -0.007675171, -0.0062828064, -0.0027103424, -0.023849487, -0.022033691, ...
[ 0.059387207, -0.019363403, 0.016036987, -0.031982422, -0.03262329, -0.011795044, -0.024139404, -0.014678955, 0.021377563, 0.038085938, -0.015380859, 0.008171082, 0.04751587, 0.0011415482, 0.0029296875, -0.005722046, -0.01109314, 0.0256958, -0.0041122437, -0.024993896, -0.0006...
http://arxiv.org/abs/2006.01563v2
http://arxiv.org/pdf/2006.01563v2
Exploring Cross-sentence Contexts for Named Entity Recognition with BERT
Jouni Luoma, Sampo Pyysalo
cs.CL
cs.CL
Named entity recognition (NER) is frequently addressed as a sequence classification task where each input consists of one sentence of text. It is nevertheless clear that useful information for the task can often be found outside of the scope of a single-sentence context. Recently proposed self-attention models such as ...
2020-12-17T16:32:55
2020-06-02T12:34:52
[ 0.010498047, -0.023513794, -0.006008148, -0.03744507, -0.021499634, 0.009002686, -0.040130615, -0.0016050339, 0.041290283, 0.022994995, -0.06378174, 0.030288696, 0.042266846, 0.011543274, 0.0053596497, 0.005077362, -0.03918457, 0.06109619, -0.007785797, -0.016830444, -0.02713...
[ 0.05618286, -0.017486572, 0.017425537, -0.027862549, -0.009666443, -0.0023231506, -0.03665161, -0.017196655, 0.02394104, 0.03817749, -0.02558899, 0.0013990402, 0.060638428, 0.0019302368, -0.00806427, 0.0018253326, -0.029846191, 0.0569458, 0.012489319, -0.011482239, -0.0064048...
http://arxiv.org/abs/2006.01592v2
http://arxiv.org/pdf/2006.01592v2
A Unified Dual-view Model for Review Summarization and Sentiment Classification with Inconsistency Loss
Hou Pong Chan, Wang Chen, Irwin King
cs.CL
cs.CL
Acquiring accurate summarization and sentiment from user reviews is an essential component of modern e-commerce platforms. Review summarization aims at generating a concise summary that describes the key opinions and sentiment of a review, while sentiment classification aims to predict a sentiment label indicating the ...
2021-02-02T08:56:42
2020-06-02T13:34:11
[ 0.036865234, -0.019882202, 0.028549194, -0.04284668, -0.05697632, 0.020828247, -0.06732178, -0.039367676, 0.027664185, 0.053985596, -0.0546875, 0.013046265, 0.01512146, 0.020019531, -0.04043579, -0.02748108, 0.00077581406, 0.032043457, 0.010948181, -0.008506775, -0.004131317,...
[ 0.039642334, -0.036499023, 0.018035889, -0.013694763, -0.050323486, -0.016677856, -0.04559326, -0.041900635, 0.03768921, 0.043518066, -0.059814453, 0.0068969727, 0.02079773, 0.02758789, -0.028259277, -0.030563354, 0.0075798035, 0.05419922, 0.044891357, -0.007987976, -0.015899...
http://arxiv.org/abs/2006.01594v1
http://arxiv.org/pdf/2006.01594v1
Training Multilingual Machine Translation by Alternately Freezing Language-Specific Encoders-Decoders
Carlos Escolano, Marta R. Costa-jussà, José A. R. Fonollosa, Mikel Artetxe
cs.CL
cs.CL, I.2.7
We propose a modular architecture of language-specific encoder-decoders that constitutes a multilingual machine translation system that can be incrementally extended to new languages without the need for retraining the existing system when adding new languages. Differently from previous works, we simultaneously train $...
2020-05-29T19:00:59
2020-05-29T19:00:59
[ -0.014152527, -0.013999939, 0.022003174, -0.015914917, -0.056610107, 0.029327393, -0.04525757, -0.018249512, 0.00573349, 0.07940674, 0.032073975, 0.0068855286, 0.050079346, -0.022583008, -0.006904602, 0.03366089, -0.008117676, 0.04663086, -0.015403748, -0.016525269, -0.022293...
[ 0.00058603287, -0.030456543, 0.0178833, -0.022918701, -0.053100586, 0.027145386, -0.09039307, -0.009407043, 0.015472412, 0.062927246, -0.009536743, 0.008239746, 0.07196045, -0.00071525574, 0.00039553642, 0.01461792, 0.006839752, 0.041809082, 0.014312744, -0.026229858, -0.0189...
http://arxiv.org/abs/2006.01603v1
http://arxiv.org/pdf/2006.01603v1
DiscSense: Automated Semantic Analysis of Discourse Markers
Damien Sileo, Tim Van de Cruys, Camille Pradel, Philippe Muller
cs.CL
cs.CL
Discourse markers ({\it by contrast}, {\it happily}, etc.) are words or phrases that are used to signal semantic and/or pragmatic relationships between clauses or sentences. Recent work has fruitfully explored the prediction of discourse markers between sentence pairs in order to learn accurate sentence representations...
2020-06-02T13:39:53
2020-06-02T13:39:53
[ 0.02519226, -0.042785645, 0.0090408325, -0.05166626, -0.052642822, 0.0068206787, -0.027313232, -0.0073547363, 0.015083313, 0.041046143, 0.014854431, -0.010597229, 0.01701355, -0.019515991, -0.03564453, 0.0062217712, -0.010925293, 0.035095215, 0.019195557, -0.015060425, 0.0036...
[ 0.064819336, -0.012924194, 0.03286743, -0.01576233, -0.016067505, -0.0058517456, -0.03451538, -0.047210693, 0.038360596, 0.027801514, 0.03262329, -0.0059013367, 0.0385437, -0.019546509, -0.012863159, 0.0013122559, -0.039855957, 0.08288574, 0.020339966, -0.021026611, 0.0095901...
http://arxiv.org/abs/2006.01715v1
http://arxiv.org/pdf/2006.01715v1
Web Document Categorization Using Naive Bayes Classifier and Latent Semantic Analysis
Alireza Saleh Sedghpour, Mohammad Reza Saleh Sedghpour
cs.CL
cs.CL
A rapid growth of web documents due to heavy use of World Wide Web necessitates efficient techniques to efficiently classify the document on the web. It is thus produced High volumes of data per second with high diversity. Automatically classification of these growing amounts of web document is One of the biggest chall...
2020-06-02T15:35:05
2020-06-02T15:35:05
[ 0.008636475, -0.002696991, -0.057861328, -0.021728516, 0.0028648376, -0.03665161, 0.007873535, 0.016311646, 0.030471802, 0.019180298, -0.03894043, -0.0002617836, 0.013557434, -0.02418518, -0.0027561188, -0.03277588, -0.0033836365, 0.057250977, 0.030807495, -0.0057144165, -0.0...
[ 0.023483276, -0.010864258, -0.06829834, -0.009719849, 0.014373779, -0.053649902, -0.02168274, 0.012893677, 0.048583984, 0.01939392, -0.043914795, -0.012992859, 0.051818848, -0.00793457, 0.02470398, -0.038970947, 0.012252808, 0.04397583, 0.025299072, -0.0082092285, -0.00772094...
http://arxiv.org/abs/2006.01912v1
http://arxiv.org/pdf/2006.01912v1
On the Predictive Power of Neural Language Models for Human Real-Time Comprehension Behavior
Ethan Gotlieb Wilcox, Jon Gauthier, Jennifer Hu, Peng Qian, Roger Levy
cs.CL
cs.CL
Human reading behavior is tuned to the statistics of natural language: the time it takes human subjects to read a word can be predicted from estimates of the word's probability in context. However, it remains an open question what computational architecture best characterizes the expectations deployed in real time by h...
2020-06-02T19:47:01
2020-06-02T19:47:01
[ -0.0056152344, -0.01663208, 0.020263672, -0.008468628, -0.08178711, 0.010475159, -0.08105469, -0.042633057, 0.032836914, 0.047943115, -0.01159668, -0.04244995, 0.03579712, -0.018997192, -0.043151855, -0.009468079, 0.007144928, 0.05581665, 0.022354126, -0.020629883, -0.0201721...
[ 0.002204895, -0.0036296844, 0.033843994, 0.036376953, -0.077941895, -0.00009083748, -0.05999756, -0.07556152, 0.0042266846, 0.037200928, -0.019119263, 0.015304565, 0.022323608, 0.001750946, -0.0022659302, -0.035186768, -0.0033626556, 0.06549072, 0.03591919, -0.024047852, -0.0...
http://arxiv.org/abs/2006.01966v1
http://arxiv.org/pdf/2006.01966v1
The Typology of Polysemy: A Multilingual Distributional Framework
Ella Rabinovich, Yang Xu, Suzanne Stevenson
cs.CL
cs.CL
Lexical semantic typology has identified important cross-linguistic generalizations about the variation and commonalities in polysemy patterns---how languages package up meanings into words. Recent computational research has enabled investigation of lexical semantics at a much larger scale, but little work has explored...
2020-06-02T22:31:40
2020-06-02T22:31:40
[ 0.010612488, -0.03050232, -0.001162529, -0.03237915, -0.035339355, 0.014175415, -0.039093018, 0.016296387, 0.003232956, 0.026733398, 0.01574707, 0.04284668, -0.025253296, -0.028900146, 0.014465332, 0.019470215, -0.0184021, 0.050201416, 0.0030136108, -0.011009216, -0.023620605...
[ 0.0058631897, -0.031707764, 0.01473999, 0.01701355, -0.02520752, 0.006942749, -0.049713135, -0.028778076, 0.019378662, 0.03540039, 0.0017662048, 0.047424316, 0.017745972, -0.015510559, 0.034210205, 0.012435913, -0.01889038, 0.044403076, 0.008506775, -0.025650024, -0.015617371...
http://arxiv.org/abs/2006.02120v1
http://arxiv.org/pdf/2006.02120v1
Towards Large-Scale Data Mining for Data-Driven Analysis of Sign Languages
Boris Mocialov, Graham Turner, Helen Hastie
cs.CL
cs.CL
Access to sign language data is far from adequate. We show that it is possible to collect the data from social networking services such as TikTok, Instagram, and YouTube by applying data filtering to enforce quality standards and by discovering patterns in the filtered data, making it easier to analyse and model. Using...
2020-06-03T09:28:17
2020-06-03T09:28:17
[ -0.01171875, -0.021591187, -0.036590576, -0.044891357, -0.030685425, 0.0178833, -0.07080078, 0.013626099, 0.04397583, 0.014465332, -0.044006348, -0.021331787, -0.024795532, -0.01184845, -0.027496338, -0.05239868, -0.010116577, 0.04711914, 0.017959595, -0.021759033, -0.0265045...
[ 0.009277344, -0.016418457, -0.011245728, -0.023361206, -0.037322998, 0.012924194, -0.07446289, -0.021209717, 0.025054932, 0.0064697266, -0.020904541, -0.026306152, -0.015731812, -0.022766113, -0.003250122, -0.04083252, -0.013679504, 0.049987793, 0.023620605, -0.025619507, -0....
http://arxiv.org/abs/2006.02144v1
http://arxiv.org/pdf/2006.02144v1
Transfer Learning for British Sign Language Modelling
Boris Mocialov, Graham Turner, Helen Hastie
cs.CL
cs.CL
Automatic speech recognition and spoken dialogue systems have made great advances through the use of deep machine learning methods. This is partly due to greater computing power but also through the large amount of data available in common languages, such as English. Conversely, research in minority languages, includin...
2020-06-03T10:13:29
2020-06-03T10:13:29
[ -0.013290405, -0.0023441315, -0.01474762, 0.005706787, -0.030838013, 0.0014925003, -0.02835083, -0.023971558, 0.028625488, 0.010887146, -0.021942139, -0.008140564, -0.025360107, -0.023483276, -0.020706177, -0.03866577, -0.01651001, 0.024963379, -0.017456055, -0.024108887, -0....
[ 0.023010254, -0.010421753, -0.0016899109, -0.0011997223, -0.030410767, -0.016311646, -0.04840088, -0.028335571, 0.031463623, 0.015586853, -0.020721436, -0.009117126, -0.0036354065, -0.028305054, -0.007827759, -0.03567505, -0.020584106, 0.05203247, -0.0005850792, -0.04397583, ...
http://arxiv.org/abs/2006.02964v1
http://arxiv.org/pdf/2006.02964v1
Personalizing Grammatical Error Correction: Adaptation to Proficiency Level and L1
Maria Nadejde, Joel Tetreault
cs.CL
cs.CL
Grammar error correction (GEC) systems have become ubiquitous in a variety of software applications, and have started to approach human-level performance for some datasets. However, very little is known about how to efficiently personalize these systems to the user's characteristics, such as their proficiency level and...
2020-06-04T15:47:29
2020-06-04T15:47:29
[ 0.014831543, 0.0021076202, -0.009132385, -0.018203735, -0.027877808, 0.012062073, -0.03451538, -0.014167786, 0.010192871, 0.05102539, 0.028915405, 0.019104004, 0.0085372925, -0.017547607, -0.023025513, 0.0037612915, -0.042816162, 0.06964111, -0.012786865, -0.06341553, 0.01172...
[ 0.022262573, -0.029418945, 0.032592773, 0.011817932, -0.056427002, 0.0020637512, -0.044525146, -0.031051636, 0.03768921, 0.06781006, -0.021697998, -0.002784729, 0.036987305, 0.019302368, -0.044158936, -0.016708374, -0.028961182, 0.060699463, 0.010040283, -0.04989624, 0.019348...
http://arxiv.org/abs/2006.03002v1
http://arxiv.org/pdf/2006.03002v1
Linguists Who Use Probabilistic Models Love Them: Quantification in Functional Distributional Semantics
Guy Emerson
cs.CL
cs.CL
Functional Distributional Semantics provides a computationally tractable framework for learning truth-conditional semantics from a corpus. Previous work in this framework has provided a probabilistic version of first-order logic, recasting quantification as Bayesian inference. In this paper, I show how the previous for...
2020-06-04T16:48:45
2020-06-04T16:48:45
[ 0.0016431808, 0.017181396, -0.024551392, -0.030380249, -0.007286072, 0.0033130646, -0.060394287, -0.026397705, 0.016479492, 0.018081665, -0.0074157715, -0.012649536, 0.010131836, -0.026123047, -0.020233154, -0.0014505386, -0.011604309, 0.05581665, 0.005054474, -0.020080566, -...
[ 0.016845703, -0.012313843, -0.01423645, -0.027114868, -0.007865906, 0.0054244995, -0.060699463, -0.013069153, 0.042907715, 0.012420654, -0.013916016, -0.022506714, 0.01133728, -0.008781433, -0.036224365, -0.007820129, -0.016571045, 0.076538086, 0.008262634, -0.02507019, -0.03...
http://arxiv.org/abs/2006.03010v1
http://arxiv.org/pdf/2006.03010v1
Syntactic Search by Example
Micah Shlain, Hillel Taub-Tabib, Shoval Sadde, Yoav Goldberg
cs.CL
cs.CL
We present a system that allows a user to search a large linguistically annotated corpus using syntactic patterns over dependency graphs. In contrast to previous attempts to this effect, we introduce a light-weight query language that does not require the user to know the details of the underlying syntactic representat...
2020-06-04T16:59:01
2020-06-04T16:59:01
[ 0.008735657, -0.027542114, 0.046691895, -0.0068359375, -0.03829956, 0.03414917, -0.034820557, -0.04385376, 0.049835205, 0.03189087, -0.03744507, 0.004928589, -0.01008606, -0.013961792, -0.03164673, -0.028656006, -0.020095825, 0.03933716, -0.014045715, -0.0031471252, 0.0037956...
[ 0.030761719, -0.04425049, 0.023590088, -0.048034668, -0.042388916, 0.02658081, -0.081848145, -0.0059318542, 0.043426514, 0.027023315, -0.008918762, -0.014091492, -0.002538681, -0.0075035095, -0.041290283, -0.032287598, -0.0020828247, 0.039154053, 0.024520874, -0.019989014, -0...
http://arxiv.org/abs/2006.03051v2
http://arxiv.org/pdf/2006.03051v2
NewB: 200,000+ Sentences for Political Bias Detection
Jerry Wei
cs.CL
cs.CL
We present the Newspaper Bias Dataset (NewB), a text corpus of more than 200,000 sentences from eleven news sources regarding Donald Trump. While previous datasets have labeled sentences as either liberal or conservative, NewB covers the political views of eleven popular media sources, capturing more nuanced political ...
2023-09-08T22:05:42
2020-06-04T18:21:50
[ 0.02394104, -0.0026378632, 0.020004272, -0.010559082, -0.032470703, -0.0075569153, -0.020004272, -0.041259766, -0.013214111, 0.0059394836, -0.035003662, 0.0109939575, -0.018966675, -0.01638794, 0.015945435, -0.01828003, -0.04208374, 0.029067993, -0.04031372, -0.030273438, -0....
[ 0.04638672, -0.025863647, 0.025527954, -0.01927185, -0.027008057, -0.017700195, -0.03768921, -0.042785645, 0.03189087, 0.03378296, -0.040100098, 0.013076782, 0.03439331, 0.012878418, 0.0039253235, -0.0065956116, -0.023956299, 0.035980225, -0.011207581, -0.038604736, -0.008720...
http://arxiv.org/abs/2006.03265v2
http://arxiv.org/pdf/2006.03265v2
Understanding Self-Attention of Self-Supervised Audio Transformers
Shu-wen Yang, Andy T. Liu, Hung-yi Lee
cs.CL
cs.CL
Self-supervised Audio Transformers (SAT) enable great success in many downstream speech applications like ASR, but how they work has not been widely explored yet. In this work, we present multiple strategies for the analysis of attention mechanisms in SAT. We categorize attentions into explainable categories, where we ...
2020-08-10T18:48:41
2020-06-05T07:23:03
[ 0.05621338, 0.0067443848, 0.0054893494, -0.05029297, 0.0008087158, -0.004043579, -0.0025634766, -0.0068359375, -0.022506714, 0.046417236, -0.011146545, 0.017074585, 0.021636963, -0.011405945, -0.0034389496, -0.004589081, -0.0027065277, 0.04458618, 0.008590698, -0.01259613, -0...
[ 0.07495117, -0.0045394897, 0.0054130554, -0.06463623, -0.010864258, -0.007827759, -0.038879395, -0.008392334, -0.014709473, 0.045166016, -0.008201599, -0.00016510487, 0.043273926, -0.010513306, -0.0011777878, -0.008049011, 0.00025057793, 0.06842041, 0.019515991, -0.020706177, ...
http://arxiv.org/abs/2006.03331v1
http://arxiv.org/pdf/2006.03331v1
ELITR Non-Native Speech Translation at IWSLT 2020
Dominik Macháček, Jonáš Kratochvíl, Sangeet Sagar, Matúš Žilinec, Ondřej Bojar, Thai-Son Nguyen, Felix Schneider, Philip Williams, Yuekun Yao
cs.CL
cs.CL
This paper is an ELITR system submission for the non-native speech translation task at IWSLT 2020. We describe systems for offline ASR, real-time ASR, and our cascaded approach to offline SLT and real-time SLT. We select our primary candidates from a pool of pre-existing systems, develop a new end-to-end general ASR sy...
2020-06-05T09:29:57
2020-06-05T09:29:57
[ 0.04525757, -0.034057617, -0.025405884, -0.088134766, -0.0022850037, -0.0053100586, -0.058258057, 0.041778564, 0.026000977, 0.0446167, -0.009399414, 0.0043525696, 0.05908203, -0.022140503, -0.010734558, 0.009490967, 0.0008454323, 0.008773804, -0.0015563965, -0.015563965, 0.03...
[ 0.046691895, -0.0184021, 0.010643005, -0.07098389, -0.021697998, 0.02142334, -0.07940674, 0.007987976, 0.0061950684, 0.04156494, -0.0073890686, -0.011680603, 0.06982422, -0.024963379, 0.0030269623, -0.003168106, 0.0073509216, 0.0390625, 0.004535675, -0.03491211, 0.010543823, ...
http://arxiv.org/abs/2006.03679v1
http://arxiv.org/pdf/2006.03679v1
Prague Dependency Treebank -- Consolidated 1.0
Jan Hajič, Eduard Bejček, Jaroslava Hlaváčová, Marie Mikulová, Milan Straka, Jan Štěpánek, Barbora Štěpánková
cs.CL
cs.CL
We present a richly annotated and genre-diversified language resource, the Prague Dependency Treebank-Consolidated 1.0 (PDT-C 1.0), the purpose of which is - as it always been the case for the family of the Prague Dependency Treebanks - to serve both as a training data for various types of NLP tasks as well as for ling...
2020-06-05T20:52:55
2020-06-05T20:52:55
[ -0.00081682205, -0.021575928, -0.064086914, -0.14978027, -0.0037937164, 0.029846191, -0.11895752, 0.07543945, 0.003042221, 0.062927246, -0.02468872, 0.016281128, -0.0040245056, -0.0011472702, -0.04928589, -0.009033203, 0.006828308, -0.031921387, -0.053985596, -0.029312134, 0....
[ 0.011497498, -0.031341553, -0.03173828, -0.06088257, -0.012458801, 0.019699097, -0.0803833, 0.029663086, 0.025222778, 0.045776367, -0.007911682, 0.018356323, 0.026977539, 0.0030345917, -0.021911621, -0.0006866455, -0.017608643, 0.042877197, -0.0074539185, -0.02708435, -0.0032...
http://arxiv.org/abs/2006.03687v1
http://arxiv.org/pdf/2006.03687v1
UDPipe at EvaLatin 2020: Contextualized Embeddings and Treebank Embeddings
Milan Straka, Jana Straková
cs.CL
cs.CL
We present our contribution to the EvaLatin shared task, which is the first evaluation campaign devoted to the evaluation of NLP tools for Latin. We submitted a system based on UDPipe 2.0, one of the winners of the CoNLL 2018 Shared Task, The 2018 Shared Task on Extrinsic Parser Evaluation and SIGMORPHON 2019 Shared Ta...
2020-06-05T21:03:35
2020-06-05T21:03:35
[ 0.052001953, 0.01663208, -0.033203125, -0.115234375, 0.022705078, 0.012779236, -0.05227661, 0.012268066, 0.019195557, 0.038146973, -0.009140015, -0.007331848, 0.0006942749, -0.017684937, -0.023284912, -0.01739502, -0.010620117, 0.046539307, 0.0041160583, -0.03665161, 0.000957...
[ 0.0047073364, -0.022537231, 0.0039367676, -0.037078857, -0.014633179, 0.036956787, -0.056793213, 0.0067863464, 0.020614624, 0.025100708, -0.018310547, -0.008277893, 0.016204834, 0.0018148422, -0.016815186, -0.013572693, -0.023117065, 0.049957275, -0.0071258545, -0.031341553, ...
http://arxiv.org/abs/2006.03866v1
http://arxiv.org/pdf/2006.03866v1
A Cross-Task Analysis of Text Span Representations
Shubham Toshniwal, Haoyue Shi, Bowen Shi, Lingyu Gao, Karen Livescu, Kevin Gimpel
cs.CL
cs.CL
Many natural language processing (NLP) tasks involve reasoning with textual spans, including question answering, entity recognition, and coreference resolution. While extensive research has focused on functional architectures for representing words and sentences, there is less work on representing arbitrary spans of te...
2020-06-06T13:37:51
2020-06-06T13:37:51
[ 0.02166748, -0.020904541, -0.017868042, 0.022384644, -0.040222168, -0.026885986, -0.0054359436, -0.0026302338, 0.035064697, 0.04711914, -0.05419922, 0.02279663, 0.038513184, 0.017410278, -0.002450943, -0.0064735413, -0.016281128, 0.059814453, -0.013946533, -0.007030487, -0.01...
[ 0.029144287, -0.03439331, 0.005306244, -0.0009717941, -0.02558899, -0.0073661804, -0.04058838, -0.0073394775, 0.03314209, 0.02859497, -0.034973145, 0.006465912, 0.04650879, 0.0018625259, 0.009918213, -0.014526367, -0.0070381165, 0.04748535, -0.0087509155, -0.023925781, -0.020...
http://arxiv.org/abs/2006.04014v1
http://arxiv.org/pdf/2006.04014v1
Medical Concept Normalization in User Generated Texts by Learning Target Concept Embeddings
Katikapalli Subramanyam Kalyan, S. Sangeetha
cs.CL
cs.CL
Medical concept normalization helps in discovering standard concepts in free-form text i.e., maps health-related mentions to standard concepts in a vocabulary. It is much beyond simple string matching and requires a deep semantic understanding of concept mentions. Recent research approach concept normalization as eithe...
2020-06-07T01:17:18
2020-06-07T01:17:18
[ 0.028945923, -0.004142761, 0.003074646, -0.005050659, -0.026062012, 0.013084412, -0.009498596, -0.04071045, 0.0021972656, 0.039154053, 0.0126953125, 0.041656494, 0.009750366, -0.0076065063, -0.0071411133, 0.0024871826, -0.0005106926, 0.015113831, -0.0010738373, -0.025344849, ...
[ 0.039642334, -0.01159668, -0.008346558, 0.0020942688, -0.010932922, -0.0126953125, -0.027526855, -0.03314209, 0.0064735413, 0.03793335, -0.015159607, 0.054351807, 0.00080633163, 0.014419556, 0.016693115, -0.018463135, -0.0037212372, 0.038726807, 0.020019531, -0.026245117, -0....
http://arxiv.org/abs/2006.04229v2
http://arxiv.org/pdf/2006.04229v2
Pre-training Polish Transformer-based Language Models at Scale
Sławomir Dadas, Michał Perełkiewicz, Rafał Poświata
cs.CL
cs.CL
Transformer-based language models are now widely used in Natural Language Processing (NLP). This statement is especially true for English language, in which many pre-trained models utilizing transformer-based architecture have been published in recent years. This has driven forward the state of the art for a variety of...
2020-06-09T12:58:43
2020-06-07T18:48:58
[ -0.022857666, -0.016113281, -0.030685425, -0.035614014, -0.038909912, 0.012207031, -0.07318115, 0.008712769, 0.016494751, 0.0435791, -0.0054130554, -0.00504303, 0.05606079, -0.028701782, 0.0041503906, 0.004508972, 0.0030670166, 0.048706055, 0.019439697, -0.031066895, 0.014694...
[ 0.015670776, -0.041107178, 0.0104599, -0.008811951, -0.033325195, 0.007419586, -0.05734253, -0.0146865845, 0.04559326, 0.04220581, -0.021697998, -0.0008149147, 0.05722046, -0.0022220612, -0.019256592, -0.013999939, -0.014602661, 0.07043457, 0.02029419, -0.018554688, -0.000449...
http://arxiv.org/abs/2006.04721v1
http://arxiv.org/pdf/2006.04721v1
Modeling Discourse Structure for Document-level Neural Machine Translation
Junxuan Chen, Xiang Li, Jiarui Zhang, Chulun Zhou, Jianwei Cui, Bin Wang, Jinsong Su
cs.CL
cs.CL
Recently, document-level neural machine translation (NMT) has become a hot topic in the community of machine translation. Despite its success, most of existing studies ignored the discourse structure information of the input document to be translated, which has shown effective in other tasks. In this paper, we propose ...
2020-06-08T16:24:03
2020-06-08T16:24:03
[ -0.011047363, -0.014472961, 0.0045433044, -0.047424316, -0.04815674, 0.034484863, -0.08654785, -0.0015745163, 0.021728516, 0.04699707, 0.039794922, -0.029647827, 0.02609253, -0.031234741, -0.019805908, -0.009361267, -0.004043579, 0.020996094, -0.032958984, -0.033111572, -0.01...
[ 0.010978699, -0.008079529, 0.026107788, -0.0037174225, -0.039642334, 0.024673462, -0.06274414, -0.031677246, 0.02029419, 0.049743652, 0.020950317, -0.0066833496, 0.07122803, -0.016967773, 0.011779785, 0.019042969, -0.003271103, 0.033935547, 0.009986877, -0.024902344, -0.00158...
http://arxiv.org/abs/2006.04781v1
http://arxiv.org/pdf/2006.04781v1
What's the Difference Between Professional Human and Machine Translation? A Blind Multi-language Study on Domain-specific MT
Lukas Fischer, Samuel Läubli
cs.CL
cs.CL
Machine translation (MT) has been shown to produce a number of errors that require human post-editing, but the extent to which professional human translation (HT) contains such errors has not yet been compared to MT. We compile pre-translated documents in which MT and HT are interleaved, and ask professional translator...
2020-06-08T17:55:14
2020-06-08T17:55:14
[ -0.006072998, -0.017837524, 0.03881836, -0.005306244, -0.07745361, 0.003894806, -0.047027588, -0.024398804, -0.024154663, 0.07519531, 0.024887085, 0.014595032, 0.045166016, -0.01600647, -0.0099487305, 0.01121521, -0.0317688, 0.026992798, -0.042419434, -0.04248047, 0.000598430...
[ 0.023513794, -0.02281189, 0.053619385, -0.003850937, -0.038085938, 0.015136719, -0.039611816, -0.04626465, 0.013999939, 0.05105591, 0.0013628006, -0.009223938, 0.07196045, -0.010437012, -0.005908966, -0.0016384125, -0.01234436, 0.044311523, 0.0067100525, -0.029144287, -0.0090...
http://arxiv.org/abs/2006.05003v1
http://arxiv.org/pdf/2006.05003v1
Universal Vector Neural Machine Translation With Effective Attention
Satish Mylapore, Ryan Quincy Paul, Joshua Yi, Robert D. Slater
cs.CL
cs.CL
Neural Machine Translation (NMT) leverages one or more trained neural networks for the translation of phrases. Sutskever introduced a sequence to sequence based encoder-decoder model which became the standard for NMT based systems. Attention mechanisms were later introduced to address the issues with the translation of...
2020-06-09T01:13:57
2020-06-09T01:13:57
[ 0.032836914, -0.0036029816, 0.022659302, -0.045715332, -0.03579712, 0.017044067, -0.058563232, -0.022521973, 0.0395813, 0.07879639, 0.03012085, -0.021606445, 0.05923462, -0.03302002, -0.039489746, 0.024261475, 0.0058403015, 0.028839111, -0.014976501, -0.0146102905, -0.0165405...
[ 0.022155762, -0.008575439, 0.037597656, -0.00390625, -0.033050537, 0.014167786, -0.070007324, -0.04751587, 0.021865845, 0.06640625, 0.011360168, -0.032073975, 0.05758667, -0.022155762, -0.02003479, -0.0076942444, 0.0012588501, 0.040924072, 0.009162903, -0.02229309, -0.0002287...
http://arxiv.org/abs/2006.05165v1
http://arxiv.org/pdf/2006.05165v1
Learning to Recover from Multi-Modality Errors for Non-Autoregressive Neural Machine Translation
Qiu Ran, Yankai Lin, Peng Li, Jie Zhou
cs.CL
cs.CL
Non-autoregressive neural machine translation (NAT) predicts the entire target sequence simultaneously and significantly accelerates inference process. However, NAT discards the dependency information in a sentence, and thus inevitably suffers from the multi-modality problem: the target tokens may be provided by differ...
2020-06-09T10:12:16
2020-06-09T10:12:16
[ -0.000852108, 0.0037288666, -0.0019683838, -0.009933472, -0.036743164, 0.012374878, -0.074279785, -0.024993896, 0.020462036, 0.06781006, 0.01058197, 0.000042676926, 0.043884277, -0.0007662773, -0.03213501, 0.006160736, -0.00178051, -0.010658264, -0.061340332, -0.033935547, -0...
[ 0.0011663437, -0.018966675, 0.0019950867, -0.022415161, -0.032318115, 0.013702393, -0.083862305, -0.012031555, 0.0074424744, 0.059387207, 0.0084991455, -0.0067214966, 0.047973633, 0.0032291412, -0.020263672, -0.007221222, -0.0009236336, 0.038085938, 0.0008468628, -0.02810669, ...
http://arxiv.org/abs/2006.05244v1
http://arxiv.org/pdf/2006.05244v1
Knowledge-Aided Open-Domain Question Answering
Mantong Zhou, Zhouxing Shi, Minlie Huang, Xiaoyan Zhu
cs.CL
cs.CL
Open-domain question answering (QA) aims to find the answer to a question from a large collection of documents.Though many models for single-document machine comprehension have achieved strong performance, there is still much room for improving open-domain QA systems since document retrieval and answer reranking are st...
2020-06-09T13:28:57
2020-06-09T13:28:57
[ 0.0066871643, -0.013130188, -0.03378296, -0.059295654, -0.03237915, -0.002248764, -0.06317139, 0.010757446, 0.035583496, 0.030014038, -0.003616333, -0.009033203, 0.0335083, -0.01210022, -0.02420044, -0.01033783, 0.010536194, -0.00006687641, -0.027496338, -0.017471313, -0.0034...
[ 0.02684021, -0.018920898, 0.0035552979, 0.008460999, -0.024887085, -0.010879517, -0.0006570816, -0.032287598, 0.049560547, 0.035949707, -0.0181427, -0.004917145, 0.01096344, 0.011238098, -0.041900635, -0.021270752, 0.002910614, 0.016723633, 0.0076560974, -0.0119018555, -0.018...
http://arxiv.org/abs/2006.05354v2
http://arxiv.org/pdf/2006.05354v2
Combination of abstractive and extractive approaches for summarization of long scientific texts
Vladislav Tretyak, Denis Stepanov
cs.CL
cs.CL
In this research work, we present a method to generate summaries of long scientific documents that uses the advantages of both extractive and abstractive approaches. Before producing a summary in an abstractive manner, we perform the extractive step, which then is used for conditioning the abstractor module. We used pr...
2020-06-12T11:25:21
2020-06-09T15:38:21
[ 0.0011768341, 0.011154175, -0.0022182465, -0.06384277, -0.03781128, 0.029754639, -0.06573486, -0.022155762, 0.044555664, 0.06530762, 0.023345947, 0.014511108, -0.023651123, -0.006160736, -0.06317139, -0.0013723373, -0.0036849976, 0.03967285, 0.006500244, -0.00096416473, -0.02...
[ 0.008079529, -0.014091492, -0.012268066, -0.048706055, -0.047576904, 0.016418457, -0.111083984, -0.016693115, 0.03326416, 0.047912598, 0.0043296814, -0.0008072853, -0.0006351471, -0.008460999, -0.026107788, -0.010246277, -0.011749268, 0.05130005, 0.030853271, -0.011108398, 0....
http://arxiv.org/abs/2006.05489v2
http://arxiv.org/pdf/2006.05489v2
Modeling Label Semantics for Predicting Emotional Reactions
Radhika Gaonkar, Heeyoung Kwon, Mohaddeseh Bastan, Niranjan Balasubramanian, Nathanael Chambers
cs.CL
cs.CL
Predicting how events induce emotions in the characters of a story is typically seen as a standard multi-label classification task, which usually treats labels as anonymous classes to predict. They ignore information that may be conveyed by the emotion labels themselves. We propose that the semantics of emotion labels ...
2020-06-28T23:47:04
2020-06-09T20:04:02
[ -0.0029087067, -0.016052246, -0.0030403137, -0.0062179565, -0.050598145, 0.0032024384, -0.027770996, -0.009635925, 0.013366699, 0.028121948, 0.0034542084, -0.005847931, 0.023895264, -0.0345459, -0.016540527, -0.0056877136, 0.023284912, 0.020095825, 0.029266357, -0.008728027, ...
[ 0.032684326, -0.030838013, -0.0050239563, -0.018707275, -0.030639648, -0.0231781, -0.030273438, -0.012710571, 0.041107178, 0.008361816, -0.0016450882, -0.009559631, 0.04171753, -0.020339966, 0.021133423, -0.0043411255, -0.01902771, 0.039489746, 0.0018901825, -0.031234741, -0....
http://arxiv.org/abs/2006.05602v1
http://arxiv.org/pdf/2006.05602v1
Adversarial Training Based Multi-Source Unsupervised Domain Adaptation for Sentiment Analysis
Yong Dai, Jian Liu, Xiancong Ren, Zenglin Xu
cs.CL
cs.CL
Multi-source unsupervised domain adaptation (MS-UDA) for sentiment analysis (SA) aims to leverage useful information in multiple source domains to help do SA in an unlabeled target domain that has no supervised information. Existing algorithms of MS-UDA either only exploit the shared features, i.e., the domain-invarian...
2020-06-10T01:41:00
2020-06-10T01:41:00
[ 0.029708862, 0.010063171, -0.012039185, -0.027496338, -0.006477356, -0.0037937164, -0.013641357, -0.02218628, 0.021194458, 0.04663086, 0.01335907, -0.026855469, 0.009674072, -0.015777588, -0.02607727, -0.0262146, 0.0021018982, 0.041107178, -0.0012359619, -0.0124435425, -0.024...
[ 0.037231445, -0.013465881, -0.003660202, -0.0052223206, -0.018371582, -0.010856628, -0.033813477, -0.029449463, 0.039123535, 0.02532959, 0.005756378, -0.018051147, 0.008598328, -0.010009766, -0.009178162, -0.016815186, -0.0003941059, 0.04522705, 0.008583069, -0.030212402, -0....
http://arxiv.org/abs/2006.05621v1
http://arxiv.org/pdf/2006.05621v1
Understanding Points of Correspondence between Sentences for Abstractive Summarization
Logan Lebanoff, John Muchovej, Franck Dernoncourt, Doo Soon Kim, Lidan Wang, Walter Chang, Fei Liu
cs.CL
cs.CL
Fusing sentences containing disparate content is a remarkable human ability that helps create informative and succinct summaries. Such a simple task for humans has remained challenging for modern abstractive summarizers, substantially restricting their applicability in real-world scenarios. In this paper, we present an...
2020-06-10T02:42:38
2020-06-10T02:42:38
[ 0.022384644, 0.00793457, 0.027282715, -0.026824951, -0.04309082, 0.030929565, -0.06738281, -0.049987793, 0.047943115, 0.06524658, -0.010910034, -0.0016107559, -0.0074310303, 0.01084137, -0.07293701, -0.0069732666, -0.012519836, 0.041290283, 0.0023097992, -0.021728516, 0.00239...
[ 0.041992188, -0.017837524, 0.0395813, -0.023422241, -0.023254395, 0.019836426, -0.057434082, -0.051757812, 0.012687683, 0.07220459, -0.003786087, 0.02619934, 0.05441284, 0.028320312, -0.017120361, -0.008407593, -0.030685425, 0.06628418, -0.0009622574, -0.021728516, -0.0056533...
http://arxiv.org/abs/2006.06110v1
http://arxiv.org/pdf/2006.06110v1
Towards Unified Dialogue System Evaluation: A Comprehensive Analysis of Current Evaluation Protocols
Sarah E. Finch, Jinho D. Choi
cs.CL
cs.CL
As conversational AI-based dialogue management has increasingly become a trending topic, the need for a standardized and reliable evaluation procedure grows even more pressing. The current state of affairs suggests various evaluation protocols to assess chat-oriented dialogue management systems, rendering it difficult ...
2020-06-10T23:29:05
2020-06-10T23:29:05
[ 0.015304565, -0.02041626, -0.032196045, -0.050872803, -0.011459351, -0.010093689, -0.04537964, -0.0054244995, 0.0008568764, 0.051116943, 0.0067443848, 0.012435913, 0.0035209656, -0.016860962, -0.03756714, -0.003768921, -0.043792725, 0.056793213, -0.049835205, -0.03881836, -0....
[ 0.012130737, -0.035949707, 0.0001399517, -0.00051021576, -0.018707275, -0.015129089, -0.026519775, -0.020095825, 0.027954102, 0.03729248, -0.0056533813, 0.0076904297, 0.014427185, -0.008705139, -0.026367188, -0.018753052, -0.034454346, 0.05227661, -0.027114868, -0.043640137, ...
http://arxiv.org/abs/2006.06202v2
http://arxiv.org/pdf/2006.06202v2
A Monolingual Approach to Contextualized Word Embeddings for Mid-Resource Languages
Pedro Javier Ortiz Suárez, Laurent Romary, Benoît Sagot
cs.CL
cs.CL
We use the multilingual OSCAR corpus, extracted from Common Crawl via language classification, filtering and cleaning, to train monolingual contextualized word embeddings (ELMo) for five mid-resource languages. We then compare the performance of OSCAR-based and Wikipedia-based ELMo embeddings for these languages on the...
2020-06-18T16:28:29
2020-06-11T05:25:18
[ 0.03793335, -0.029907227, -0.0028705597, -0.041107178, -0.040740967, 0.0119018555, -0.054382324, 0.008583069, -0.0044670105, 0.057434082, -0.03277588, 0.025878906, 0.02897644, 0.003019333, 0.005592346, -0.0048179626, -0.00919342, 0.043426514, -0.0012798309, -0.017745972, -0.0...
[ 0.05697632, -0.025787354, 0.012283325, -0.03161621, -0.008674622, -0.0059394836, -0.039245605, -0.01020813, -0.026184082, 0.05441284, -0.016525269, 0.021057129, 0.033233643, 0.0107421875, 0.016723633, -0.023284912, -0.024734497, 0.07232666, 0.010398865, -0.016998291, -0.01681...
http://arxiv.org/abs/2006.06226v1
http://arxiv.org/pdf/2006.06226v1
Discrete Latent Variable Representations for Low-Resource Text Classification
Shuning Jin, Sam Wiseman, Karl Stratos, Karen Livescu
cs.CL
cs.CL
While much work on deep latent variable models of text uses continuous latent variables, discrete latent variables are interesting because they are more interpretable and typically more space efficient. We consider several approaches to learning discrete latent variable models for text in the case where exact marginali...
2020-06-11T06:55:13
2020-06-11T06:55:13
[ 0.006904602, -0.022140503, -0.048675537, -0.026626587, -0.036193848, -0.01852417, -0.03668213, 0.0023422241, 0.027359009, 0.024429321, -0.057159424, 0.024230957, -0.0053596497, -0.004901886, -0.0066719055, -0.041900635, -0.0056381226, 0.05871582, -0.011634827, -0.016448975, -...
[ 0.0098724365, -0.023910522, -0.012008667, -0.0032081604, -0.051635742, -0.0020751953, -0.06915283, -0.0087890625, 0.020599365, 0.024047852, -0.037506104, 0.02758789, 0.029388428, 0.0047187805, 0.014274597, -0.035827637, -0.013252258, 0.05166626, -0.0048599243, -0.020523071, -...
http://arxiv.org/abs/2006.06259v1
http://arxiv.org/pdf/2006.06259v1
Augmenting Data for Sarcasm Detection with Unlabeled Conversation Context
Hankyol Lee, Youngjae Yu, Gunhee Kim
cs.CL
cs.CL
We present a novel data augmentation technique, CRA (Contextual Response Augmentation), which utilizes conversational context to generate meaningful samples for training. We also mitigate the issues regarding unbalanced context lengths by changing the input-output format of the model such that it can deal with varying ...
2020-06-11T09:00:11
2020-06-11T09:00:11
[ 0.015838623, 0.004398346, -0.04107666, -0.03048706, -0.007850647, -0.006679535, -0.052764893, -0.032287598, 0.064086914, 0.003206253, -0.035064697, -0.037750244, 0.047576904, -0.020568848, -0.0038833618, -0.02810669, -0.016799927, 0.08068848, -0.004020691, -0.01927185, -0.024...
[ 0.04916382, -0.0053253174, -0.013374329, -0.026000977, -0.018997192, -0.023025513, -0.049316406, -0.0064582825, 0.012290955, 0.03857422, -0.028793335, -0.0033283234, 0.057128906, -0.017593384, 0.03062439, 0.018844604, -0.017059326, 0.08258057, -0.010620117, -0.041229248, 0.01...
http://arxiv.org/abs/2006.06264v2
http://arxiv.org/pdf/2006.06264v2
Tangled up in BLEU: Reevaluating the Evaluation of Automatic Machine Translation Evaluation Metrics
Nitika Mathur, Timothy Baldwin, Trevor Cohn
cs.CL
cs.CL
Automatic metrics are fundamental for the development and evaluation of machine translation systems. Judging whether, and to what extent, automatic metrics concur with the gold standard of human evaluation is not a straightforward problem. We show that current methods for judging metrics are highly sensitive to the tra...
2020-06-12T04:35:41
2020-06-11T09:12:53
[ -0.048461914, -0.020111084, 0.0078125, 0.026046753, -0.04473877, 0.01386261, -0.021438599, -0.029830933, 0.017608643, 0.05947876, -0.0016469955, -0.011459351, 0.033325195, -0.018051147, -0.03652954, -0.0064811707, -0.011558533, 0.019882202, -0.0385437, -0.056121826, -0.031646...
[ -0.029403687, -0.039398193, 0.027999878, 0.037506104, -0.061065674, 0.008506775, -0.051452637, -0.046783447, 0.013587952, 0.0748291, 0.0053901672, 0.0063209534, 0.061431885, -0.0012769699, -0.034942627, -0.013511658, -0.0075569153, 0.023803711, -0.019088745, -0.051849365, 0.0...
http://arxiv.org/abs/2006.06341v2
http://arxiv.org/pdf/2006.06341v2
Provenance for Linguistic Corpora Through Nanopublications
Timo Lek, Anna de Groot, Tobias Kuhn, Roser Morante
cs.CL
cs.CL
Research in Computational Linguistics is dependent on text corpora for training and testing new tools and methodologies. While there exists a plethora of annotated linguistic information, these corpora are often not interoperable without significant manual work. Moreover, these annotations might have evolved into diffe...
2020-11-02T07:29:09
2020-06-11T11:30:30
[ -0.0054016113, -0.038879395, -0.021774292, -0.04525757, -0.047180176, 0.031677246, -0.06390381, 0.022628784, 0.031158447, 0.022399902, -0.008132935, -0.023361206, 0.026412964, -0.02696228, -0.010971069, 0.000061929226, -0.012664795, 0.02885437, -0.002281189, -0.0141067505, -0...
[ 0.022064209, -0.029571533, -0.012557983, -0.06524658, -0.016098022, 0.01676941, -0.0715332, 0.007843018, 0.043304443, 0.021957397, -0.0037879944, -0.0009851456, 0.016815186, -0.007774353, -0.008583069, -0.02078247, -0.021835327, 0.043304443, -0.0003209114, -0.022827148, -0.02...
http://arxiv.org/abs/2006.06402v2
http://arxiv.org/pdf/2006.06402v2
CoSDA-ML: Multi-Lingual Code-Switching Data Augmentation for Zero-Shot Cross-Lingual NLP
Libo Qin, Minheng Ni, Yue Zhang, Wanxiang Che
cs.CL
cs.CL
Multi-lingual contextualized embeddings, such as multilingual-BERT (mBERT), have shown success in a variety of zero-shot cross-lingual tasks. However, these models are limited by having inconsistent contextualized representations of subwords across different languages. Existing work addresses this issue by bilingual pr...
2020-07-13T04:59:04
2020-06-11T13:15:59
[ 0.018096924, -0.040252686, 0.0029277802, -0.020019531, -0.03781128, 0.0013532639, -0.02633667, 0.006713867, 0.04510498, 0.03201294, -0.035095215, 0.004436493, 0.036956787, -0.017532349, 0.005039215, 0.014160156, -0.008895874, 0.07318115, 0.0068244934, -0.016220093, -0.0243377...
[ 0.05633545, -0.016723633, 0.026306152, -0.014915466, -0.02432251, -0.00390625, -0.043304443, -0.023025513, 0.010932922, 0.0519104, -0.017913818, 0.0043563843, 0.0602417, -0.005001068, 0.01966858, -0.0018548965, -0.016296387, 0.066101074, 0.010787964, -0.025558472, -0.02444458...
http://arxiv.org/abs/2006.06478v2
http://arxiv.org/pdf/2006.06478v2
Multi-hop Reading Comprehension across Documents with Path-based Graph Convolutional Network
Zeyun Tang, Yongliang Shen, Xinyin Ma, Wei Xu, Jiale Yu, Weiming Lu
cs.CL
cs.CL
Multi-hop reading comprehension across multiple documents attracts much attention recently. In this paper, we propose a novel approach to tackle this multi-hop reading comprehension problem. Inspired by human reasoning processing, we construct a path-based reasoning graph from supporting documents. This graph can combi...
2020-06-12T04:04:05
2020-06-11T14:43:34
[ -0.027648926, -0.044189453, -0.010269165, -0.031677246, -0.08074951, 0.014274597, -0.081848145, 0.019821167, 0.03994751, 0.027435303, -0.041503906, -0.010696411, 0.026382446, -0.011367798, -0.0040397644, -0.018417358, -0.0028038025, 0.05532837, 0.0138549805, -0.01448822, -0.0...
[ -0.018569946, -0.047607422, 0.0075950623, -0.017608643, -0.06524658, 0.011001587, -0.06585693, -0.00054454803, 0.028930664, 0.03564453, -0.022384644, -0.0018978119, 0.022338867, 0.005332947, -0.016311646, -0.025787354, -0.0037403107, 0.028808594, 0.014434814, -0.017425537, -0...
http://arxiv.org/abs/2006.07015v1
http://arxiv.org/pdf/2006.07015v1
Speaker Sensitive Response Evaluation Model
JinYeong Bak, Alice Oh
cs.CL
cs.CL
Automatic evaluation of open-domain dialogue response generation is very challenging because there are many appropriate responses for a given context. Existing evaluation models merely compare the generated response with the ground truth response and rate many of the appropriate responses as inappropriate if they devia...
2020-06-12T08:59:10
2020-06-12T08:59:10
[ 0.020828247, 0.030899048, 0.011497498, 0.021850586, -0.0019483566, -0.00573349, 0.0033512115, -0.04046631, 0.009086609, 0.06640625, 0.019866943, -0.02142334, -0.013023376, -0.002779007, -0.083496094, -0.0070343018, -0.022506714, 0.028060913, -0.029968262, -0.025985718, 0.0148...
[ 0.024169922, -0.007446289, -0.015090942, -0.011985779, -0.009986877, -0.016586304, -0.021636963, -0.022537231, 0.000603199, 0.034698486, -0.006591797, 0.023452759, -0.0059547424, 0.0010995865, -0.034088135, -0.031036377, -0.02520752, 0.0513916, -0.017990112, -0.03427124, 0.00...
http://arxiv.org/abs/2006.07235v2
http://arxiv.org/pdf/2006.07235v2
SemEval-2020 Task 12: Multilingual Offensive Language Identification in Social Media (OffensEval 2020)
Marcos Zampieri, Preslav Nakov, Sara Rosenthal, Pepa Atanasova, Georgi Karadzhov, Hamdy Mubarak, Leon Derczynski, Zeses Pitenis, Çağrı Çöltekin
cs.CL
cs.CL, 68T50, 68T07, I.2.7
We present the results and main findings of SemEval-2020 Task 12 on Multilingual Offensive Language Identification in Social Media (OffensEval 2020). The task involves three subtasks corresponding to the hierarchical taxonomy of the OLID schema (Zampieri et al., 2019a) from OffensEval 2019. The task featured five langu...
2020-09-30T15:46:44
2020-06-12T14:39:40
[ 0.047088623, 0.009941101, -0.027801514, -0.03652954, 0.015464783, -0.015563965, -0.026367188, -0.017852783, 0.01838684, 0.03527832, 0.004005432, 0.010986328, 0.00076150894, -0.009773254, -0.016571045, 0.001832962, -0.016845703, 0.07861328, 0.044830322, -0.008995056, -0.015342...
[ 0.06414795, 0.009475708, -0.026260376, -0.049591064, 0.033599854, -0.02558899, -0.040222168, -0.017562866, 0.017059326, 0.018417358, 0.0028190613, 0.016540527, 0.0045547485, -0.017318726, 0.015777588, -0.010284424, 0.0059776306, 0.05682373, 0.053100586, -0.007713318, -0.02661...
http://arxiv.org/abs/2006.07264v1
http://arxiv.org/pdf/2006.07264v1
Low-resource Languages: A Review of Past Work and Future Challenges
Alexandre Magueresse, Vincent Carles, Evan Heetderks
cs.CL
cs.CL, I.2.7
A current problem in NLP is massaging and processing low-resource languages which lack useful training attributes such as supervised data, number of native speakers or experts, etc. This review paper concisely summarizes previous groundbreaking achievements made towards resolving this problem, and analyzes potential im...
2020-06-12T15:21:57
2020-06-12T15:21:57
[ -0.011627197, -0.05041504, -0.02229309, -0.03201294, -0.055908203, 0.004131317, -0.057891846, 0.021347046, -0.013198853, 0.049194336, 0.023040771, -0.013656616, 0.02104187, -0.039611816, -0.022369385, 0.0059318542, -0.016815186, 0.05432129, 0.009986877, -0.019256592, -0.00240...
[ 0.026885986, -0.04611206, 0.011482239, -0.011451721, -0.039916992, 0.011749268, -0.0501709, -0.008979797, 0.032440186, 0.024917603, -0.009307861, 0.001830101, 0.026184082, 0.00012153387, 0.0046691895, -0.012931824, -0.0067367554, 0.0680542, 0.003227234, -0.03062439, -0.028137...
http://arxiv.org/abs/2006.07296v6
http://arxiv.org/pdf/2006.07296v6
Information Extraction of Clinical Trial Eligibility Criteria
Yitong Tseo, M. I. Salkola, Ahmed Mohamed, Anuj Kumar, Freddy Abnousi
cs.CL
cs.CL
Clinical trials predicate subject eligibility on a diversity of criteria ranging from patient demographics to food allergies. Trials post their requirements as semantically complex, unstructured free-text. Formalizing trial criteria to a computer-interpretable syntax would facilitate eligibility determination. In this ...
2020-07-28T17:50:42
2020-06-12T16:25:45
[ 0.013336182, 0.0035381317, -0.026916504, -0.03918457, -0.012008667, 0.023864746, -0.02885437, -0.027374268, -0.0044898987, 0.0073280334, -0.024429321, 0.009757996, -0.044433594, -0.016830444, -0.008132935, -0.033355713, -0.029968262, 0.028900146, -0.044128418, -0.03137207, 0....
[ 0.035308838, -0.017288208, -0.019943237, -0.07043457, -0.0018548965, 0.01133728, -0.05984497, -0.018463135, -0.00036215782, 0.016738892, -0.018600464, 0.017745972, -0.0005145073, -0.0057792664, 0.02571106, -0.033081055, -0.042541504, 0.06323242, -0.01687622, -0.055999756, 0.0...
http://arxiv.org/abs/2006.07425v2
http://arxiv.org/pdf/2006.07425v2
Measuring Forecasting Skill from Text
Shi Zong, Alan Ritter, Eduard Hovy
cs.CL
cs.CL
People vary in their ability to make accurate predictions about the future. Prior studies have shown that some individuals can predict the outcome of future events with consistently better accuracy. This leads to a natural question: what makes some forecasters better than others? In this paper we explore connections be...
2020-06-16T16:09:30
2020-06-12T19:04:10
[ -0.041870117, -0.014259338, -0.0010099411, 0.018798828, -0.051513672, 0.013504028, -0.011520386, -0.018341064, 0.03152466, 0.048461914, -0.035308838, 0.008422852, -0.029525757, 0.012084961, -0.064086914, -0.044006348, 0.014968872, 0.004497528, 0.0006132126, -0.006061554, 0.01...
[ 0.027893066, -0.023849487, -0.017837524, -0.02935791, -0.025939941, -0.009155273, -0.057739258, -0.013252258, 0.029312134, 0.04345703, 0.0037879944, -0.024841309, 0.020751953, 0.006839752, -0.054229736, -0.03768921, 0.008132935, 0.028961182, 0.0463562, -0.014442444, -0.002973...
http://arxiv.org/abs/2006.07667v1
http://arxiv.org/pdf/2006.07667v1
Words ranking and Hirsch index for identifying the core of the hapaxes in political texts
Valerio Ficcadenti, Roy Cerqueti, Marcel Ausloos, Gurjeet Dhesi
cs.CL
cs.CL
This paper deals with a quantitative analysis of the content of official political speeches. We study a set of about one thousand talks pronounced by the US Presidents, ranging from Washington to Trump. In particular, we search for the relevance of the rare words, i.e. those said only once in each speech -- the so-call...
2020-06-13T15:48:15
2020-06-13T15:48:15
[ -0.01676941, -0.0030212402, 0.013008118, 0.031234741, -0.019683838, -0.01902771, 0.028457642, -0.0423584, 0.013267517, 0.03137207, -0.032348633, 0.033111572, -0.018051147, -0.0022964478, -0.01890564, -0.023468018, -0.028839111, 0.019378662, -0.0134887695, -0.01876831, 0.01809...
[ 0.011016846, -0.023361206, -0.015914917, 0.015533447, -0.024291992, -0.046203613, 0.006877899, -0.021530151, -0.0063209534, 0.018692017, -0.022354126, 0.0309906, -0.006767273, 0.008270264, -0.0038414001, -0.041503906, -0.038635254, 0.047332764, 0.011360168, -0.04336548, -0.00...
http://arxiv.org/abs/2006.07804v1
http://arxiv.org/pdf/2006.07804v1
Vietnamese Word Segmentation with SVM: Ambiguity Reduction and Suffix Capture
Duc-Vu Nguyen, Dang Van Thin, Kiet Van Nguyen, Ngan Luu-Thuy Nguyen
cs.CL
cs.CL
In this paper, we approach Vietnamese word segmentation as a binary classification by using the Support Vector Machine classifier. We inherit features from prior works such as n-gram of syllables, n-gram of syllable types, and checking conjunction of adjacent syllables in the dictionary. We propose two novel ways to fe...
2020-06-14T05:19:46
2020-06-14T05:19:46
[ -0.013664246, -0.019638062, -0.0040283203, -0.0013837814, -0.053771973, 0.011550903, -0.049926758, -0.023223877, 0.013000488, 0.036987305, -0.020904541, -0.022842407, -0.001964569, -0.017303467, -0.036193848, -0.027618408, -0.0206604, 0.028259277, -0.011550903, -0.008087158, ...
[ -0.0050239563, -0.024414062, 0.0013856888, 0.0065460205, -0.03753662, 0.0053100586, -0.055725098, -0.023742676, 0.018005371, 0.035095215, -0.016464233, -0.02230835, -0.00907135, -0.00095558167, -0.02268982, -0.044921875, -0.013374329, 0.049102783, 0.015731812, -0.014801025, 0...
http://arxiv.org/abs/2006.07890v1
http://arxiv.org/pdf/2006.07890v1
FinEst BERT and CroSloEngual BERT: less is more in multilingual models
Matej Ulčar, Marko Robnik-Šikonja
cs.CL
cs.CL
Large pretrained masked language models have become state-of-the-art solutions for many NLP problems. The research has been mostly focused on English language, though. While massively multilingual models exist, studies have shown that monolingual models produce much better results. We train two trilingual BERT-like mod...
2020-06-14T12:54:01
2020-06-14T12:54:01
[ -0.02684021, -0.0619812, -0.01348114, -0.021575928, -0.06878662, 0.0098724365, -0.072631836, 0.029281616, -0.015014648, 0.070007324, 0.0055351257, 0.013549805, 0.045715332, -0.007095337, 0.00579834, 0.005508423, -0.017730713, 0.020736694, -0.013153076, -0.032348633, 0.0074005...
[ 0.026611328, -0.0390625, 0.0075531006, -0.03704834, -0.04043579, 0.017074585, -0.101989746, -0.0026130676, -0.01600647, 0.044677734, -0.0020561218, 0.0075302124, 0.048614502, -0.007858276, 0.013458252, -0.0079574585, -0.026321411, 0.05630493, -0.0029315948, -0.031036377, -0.0...
http://arxiv.org/abs/2006.08097v2
http://arxiv.org/pdf/2006.08097v2
FinBERT: A Pretrained Language Model for Financial Communications
Yi Yang, Mark Christopher Siy UY, Allen Huang
cs.CL
cs.CL
Contextual pretrained language models, such as BERT (Devlin et al., 2019), have made significant breakthrough in various NLP tasks by training on large scale of unlabeled text re-sources.Financial sector also accumulates large amount of financial communication text.However, there is no pretrained finance specific langu...
2020-07-09T02:50:04
2020-06-15T02:51:06
[ -0.015281677, -0.022033691, -0.03717041, -0.04031372, -0.034057617, 0.0052757263, -0.0647583, 0.009414673, 0.010612488, 0.05368042, 0.0024375916, -0.018707275, 0.04208374, -0.010696411, -0.04071045, -0.006526947, -0.025253296, 0.055541992, 0.045776367, -0.021591187, -0.004505...
[ 0.03149414, -0.012802124, -0.022415161, -0.027557373, -0.021331787, -0.008399963, -0.059448242, -0.021224976, 0.050354004, 0.042419434, -0.038024902, -0.015625, 0.05505371, 0.022354126, -0.035339355, -0.009590149, -0.040771484, 0.06402588, 0.02204895, -0.019577026, -0.0196990...
http://arxiv.org/abs/2006.08101v1
http://arxiv.org/pdf/2006.08101v1
Evidence-Aware Inferential Text Generation with Vector Quantised Variational AutoEncoder
Daya Guo, Duyu Tang, Nan Duan, Jian Yin, Daxin Jiang, Ming Zhou
cs.CL
cs.CL
Generating inferential texts about an event in different perspectives requires reasoning over different contexts that the event occurs. Existing works usually ignore the context that is not explicitly provided, resulting in a context-independent semantic representation that struggles to support the generation. To addre...
2020-06-15T02:59:52
2020-06-15T02:59:52
[ 0.0069732666, -0.031982422, -0.026824951, -0.01651001, -0.07836914, 0.0037784576, -0.062805176, -0.005378723, 0.021270752, 0.044677734, 0.010559082, 0.0112838745, 0.019836426, -0.01739502, -0.014701843, 0.01007843, -0.0071144104, 0.046020508, -0.026153564, -0.022521973, -0.01...
[ 0.016693115, -0.01953125, -0.005004883, -0.032806396, -0.032470703, 0.0033435822, -0.056274414, -0.017593384, 0.019744873, 0.033294678, 0.0037956238, 0.01360321, 0.05218506, -0.005695343, 0.004070282, 0.009048462, -0.008399963, 0.030166626, 0.0031547546, -0.018661499, -0.0104...
http://arxiv.org/abs/2006.08185v1
http://arxiv.org/pdf/2006.08185v1
Extracting N-ary Cross-sentence Relations using Constrained Subsequence Kernel
Sachin Pawar, Pushpak Bhattacharyya, Girish K. Palshikar
cs.CL
cs.CL
Most of the past work in relation extraction deals with relations occurring within a sentence and having only two entity arguments. We propose a new formulation of the relation extraction task where the relations are more general than intra-sentence relations in the sense that they may span multiple sentences and may h...
2020-06-15T07:23:58
2020-06-15T07:23:58
[ -0.020202637, -0.023025513, -0.04257202, -0.03149414, -0.032684326, 0.00349617, -0.0418396, 0.012924194, 0.042236328, 0.013717651, -0.018585205, -0.015098572, 0.006626129, -0.019485474, -0.021759033, -0.019851685, -0.033294678, 0.047576904, -0.03237915, -0.004676819, -0.02378...
[ 0.049682617, -0.023544312, -0.007965088, -0.020767212, -0.019378662, -0.01184082, -0.0579834, -0.012893677, 0.06689453, 0.003921509, -0.018478394, -0.015434265, 0.0015220642, -0.009819031, -0.01449585, -0.027404785, -0.018707275, 0.06713867, 0.015792847, -0.0209198, 0.0016870...
http://arxiv.org/abs/2006.08297v1
http://arxiv.org/pdf/2006.08297v1
Fine-grained Human Evaluation of Transformer and Recurrent Approaches to Neural Machine Translation for English-to-Chinese
Yuying Ye, Antonio Toral
cs.CL
cs.CL
This research presents a fine-grained human evaluation to compare the Transformer and recurrent approaches to neural machine translation (MT), on the translation direction English-to-Chinese. To this end, we develop an error taxonomy compliant with the Multidimensional Quality Metrics (MQM) framework that is customised...
2020-06-15T11:47:00
2020-06-15T11:47:00
[ 0.021392822, 0.013381958, 0.043273926, 0.020843506, -0.021362305, 0.003736496, 0.00011098385, -0.035125732, 0.004890442, 0.05291748, 0.011009216, -0.0066490173, 0.03289795, -0.031341553, -0.033081055, 0.0066566467, -0.010124207, 0.041809082, -0.038604736, -0.04232788, -0.0417...
[ 0.00957489, -0.00356102, 0.028564453, 0.02458191, -0.029449463, 0.0058555603, -0.037475586, -0.042938232, 0.011207581, 0.056671143, 0.01586914, -0.0016002655, 0.050811768, -0.01966858, -0.013542175, 0.009635925, -0.014694214, 0.048797607, -0.004722595, -0.041229248, -0.002349...
http://arxiv.org/abs/2006.08332v1
http://arxiv.org/pdf/2006.08332v1
An Augmented Translation Technique for low Resource language pair: Sanskrit to Hindi translation
Rashi Kumar, Piyush Jha, Vineet Sahula
cs.CL
cs.CL, I.2.7
Neural Machine Translation (NMT) is an ongoing technique for Machine Translation (MT) using enormous artificial neural network. It has exhibited promising outcomes and has shown incredible potential in solving challenging machine translation exercises. One such exercise is the best approach to furnish great MT to langu...
2020-06-09T17:01:55
2020-06-09T17:01:55
[ -0.0418396, -0.018493652, 0.011810303, -0.009788513, -0.05053711, 0.024139404, -0.04360962, -0.006877899, 0.029937744, 0.030883789, 0.037506104, -0.009254456, -0.018249512, -0.042236328, -0.04840088, -0.010559082, -0.013679504, 0.038757324, -0.0036754608, -0.031051636, -0.002...
[ -0.010520935, -0.008621216, 0.012809753, 0.005378723, -0.03173828, 0.0135650635, -0.058654785, -0.03289795, -0.0035171509, 0.04067993, 0.019836426, -0.012634277, 0.015068054, -0.021377563, -0.016952515, -0.009460449, -0.021957397, 0.06173706, 0.031677246, -0.03555298, 0.01545...
http://arxiv.org/abs/2006.08337v2
http://arxiv.org/pdf/2006.08337v2
Open-Domain Question Answering with Pre-Constructed Question Spaces
Jinfeng Xiao, Lidan Wang, Franck Dernoncourt, Trung Bui, Tong Sun, Jiawei Han
cs.CL
cs.CL
Open-domain question answering aims at solving the task of locating the answers to user-generated questions in massive collections of documents. There are two families of solutions available: retriever-readers, and knowledge-graph-based approaches. A retriever-reader usually first uses information retrieval methods lik...
2020-10-23T02:10:07
2020-06-02T04:31:09
[ 0.01876831, -0.022476196, 0.0074806213, -0.047912598, -0.03378296, 0.017837524, -0.03677368, -0.00818634, 0.02180481, -0.0011720657, -0.018737793, -0.0019893646, -0.028244019, -0.022857666, -0.010185242, -0.038085938, 0.007785797, -0.021530151, -0.040985107, -0.030288696, -0....
[ 0.028701782, -0.011131287, 0.0065689087, -0.000187397, -0.0211792, -0.0007867813, -0.027496338, -0.032836914, 0.04220581, 0.021530151, -0.017333984, 0.0002501011, 0.005168915, 0.008865356, -0.015075684, -0.02545166, 0.009231567, 0.020828247, 0.0016031265, -0.017227173, -0.006...
http://arxiv.org/abs/2006.08342v2
http://arxiv.org/pdf/2006.08342v2
Subjective Question Answering: Deciphering the inner workings of Transformers in the realm of subjectivity
Lukas Muttenthaler
cs.CL
cs.CL
Understanding subjectivity demands reasoning skills beyond the realm of common knowledge. It requires a machine learning model to process sentiment and to perform opinion mining. In this work, I've exploited a recently released dataset for span-selection Question Answering, namely SubjQA. SubjQA is the first QA dataset...
2020-10-14T07:47:52
2020-06-02T13:48:14
[ 0.03387451, -0.0178833, 0.008323669, -0.0062217712, -0.027816772, -0.011383057, 0.026504517, -0.019989014, 0.035888672, 0.028427124, -0.0357666, 0.0076179504, -0.0033988953, -0.005332947, -0.051696777, -0.018463135, 0.0064582825, 0.046325684, 0.008491516, -0.010482788, -0.032...
[ 0.033050537, -0.028640747, 0.023544312, 0.0357666, -0.03744507, -0.026535034, 0.040283203, -0.040924072, 0.049316406, 0.01625061, -0.04925537, -0.014907837, 0.0034236908, 0.0104522705, -0.03540039, -0.052947998, -0.019012451, 0.046569824, -0.0010623932, -0.016296387, -0.01791...
http://arxiv.org/abs/2006.08387v2
http://arxiv.org/pdf/2006.08387v2
Catplayinginthesnow: Impact of Prior Segmentation on a Model of Visually Grounded Speech
William N. Havard, Jean-Pierre Chevrot, Laurent Besacier
cs.CL
cs.CL
The language acquisition literature shows that children do not build their lexicon by segmenting the spoken input into phonemes and then building up words from them, but rather adopt a top-down approach and start by segmenting word-like units and then break them down into smaller units. This suggests that the ideal way...
2020-10-20T13:15:59
2020-06-15T13:20:13
[ 0.034301758, 0.009262085, -0.003929138, -0.04925537, -0.02861023, -0.0019426346, -0.04168701, -0.01272583, -0.007129669, 0.041625977, -0.010940552, 0.00038051605, 0.014389038, -0.03125, -0.012863159, -0.027008057, 0.0046958923, 0.0715332, 0.028930664, -0.017623901, -0.0096435...
[ 0.012435913, -0.017089844, 0.008384705, 0.011268616, -0.054260254, 0.004184723, -0.03225708, -0.009849548, -0.01637268, 0.040924072, -0.004257202, 0.016693115, -0.0020771027, -0.011688232, 0.008613586, -0.032592773, -0.013145447, 0.050628662, 0.0040626526, -0.020324707, -0.00...
http://arxiv.org/abs/2006.08748v1
http://arxiv.org/pdf/2006.08748v1
DynE: Dynamic Ensemble Decoding for Multi-Document Summarization
Chris Hokamp, Demian Gholipour Ghalandari, Nghia The Pham, John Glover
cs.CL
cs.CL
Sequence-to-sequence (s2s) models are the basis for extensive work in natural language processing. However, some applications, such as multi-document summarization, multi-modal machine translation, and the automatic post-editing of machine translation, require mapping a set of multiple distinct inputs into a single out...
2020-06-15T20:40:06
2020-06-15T20:40:06
[ -0.00006479025, -0.013343811, -0.039916992, -0.08050537, -0.03439331, 0.014221191, -0.09625244, 0.0005455017, 0.020004272, 0.06744385, -0.012481689, -0.021347046, 0.04510498, -0.0058937073, -0.04928589, -0.009513855, 0.008918762, 0.037139893, 0.018753052, -0.011489868, -0.030...
[ 0.025421143, -0.014503479, 0.0060424805, -0.05050659, -0.032104492, 0.015533447, -0.09136963, -0.013587952, 0.02243042, 0.061920166, 0.018554688, -0.0026454926, 0.047180176, -0.009757996, -0.020339966, 0.01272583, 0.009262085, 0.04498291, 0.0030765533, -0.010879517, -0.011199...
http://arxiv.org/abs/2006.08881v1
http://arxiv.org/pdf/2006.08881v1
Scalable Cross Lingual Pivots to Model Pronoun Gender for Translation
Kellie Webster, Emily Pitler
cs.CL
cs.CL
Machine translation systems with inadequate document understanding can make errors when translating dropped or neutral pronouns into languages with gendered pronouns (e.g., English). Predicting the underlying gender of these pronouns is difficult since it is not marked textually and must instead be inferred from corefe...
2020-06-16T02:41:46
2020-06-16T02:41:46
[ -0.04119873, -0.03778076, 0.0055885315, -0.07067871, -0.05166626, 0.04324341, -0.07788086, 0.019821167, 0.071899414, 0.050476074, 0.0042762756, -0.017166138, 0.06915283, -0.017684937, -0.031188965, 0.019515991, -0.016174316, 0.024032593, -0.022537231, -0.044067383, -0.0155029...
[ -0.016616821, -0.03564453, 0.032348633, -0.009757996, -0.050872803, 0.029586792, -0.08508301, -0.030090332, 0.0592041, 0.053344727, -0.006679535, -0.019470215, 0.0826416, -0.0036830902, -0.02218628, 0.01335144, -0.032928467, 0.049743652, -0.004837036, -0.048614502, -0.0125274...
http://arxiv.org/abs/2006.09035v1
http://arxiv.org/pdf/2006.09035v1
The SPPD System for Schema Guided Dialogue State Tracking Challenge
Miao Li, Haoqi Xiong, Yunbo Cao
cs.CL
cs.CL
This paper introduces one of our group's work on the Dialog System Technology Challenges 8 (DSTC8), the SPPD system for Schema Guided dialogue state tracking challenge. This challenge, named as Track 4 in DSTC8, provides a brand new and challenging dataset for developing scalable multi-domain dialogue state tracking al...
2020-06-16T09:57:40
2020-06-16T09:57:40
[ 0.010368347, -0.05935669, -0.021911621, -0.06451416, -0.047180176, 0.009735107, -0.047943115, 0.027374268, 0.0259552, 0.005657196, 0.021636963, -0.012519836, -0.029632568, -0.046203613, -0.010810852, -0.019515991, 0.0012292862, -0.0018062592, -0.017501831, -0.008354187, -0.01...
[ 0.04031372, -0.044525146, -0.02154541, -0.065979004, -0.015914917, -0.020004272, -0.04937744, 0.01461792, 0.039611816, 0.010093689, 0.00027441978, -0.0031013489, 0.018844604, -0.02684021, 0.010269165, -0.0018482208, -0.017745972, 0.013442993, -0.008674622, -0.032440186, 0.008...
http://arxiv.org/abs/2006.09109v2
http://arxiv.org/pdf/2006.09109v2
How to Probe Sentence Embeddings in Low-Resource Languages: On Structural Design Choices for Probing Task Evaluation
Steffen Eger, Johannes Daxenberger, Iryna Gurevych
cs.CL
cs.CL
Sentence encoders map sentences to real valued vectors for use in downstream applications. To peek into these representations - e.g., to increase interpretability of their results - probing tasks have been designed which query them for linguistic knowledge. However, designing probing tasks for lesser-resourced language...
2020-10-28T12:38:37
2020-06-16T12:37:50
[ -0.0024642944, -0.04711914, 0.025497437, 0.015991211, -0.072631836, 0.009185791, -0.028305054, -0.01637268, 0.010139465, 0.0050735474, -0.019134521, 0.0030326843, -0.009498596, -0.023483276, -0.01687622, -0.02999878, -0.020874023, 0.042236328, -0.03604126, -0.024993896, -0.03...
[ 0.0016756058, -0.04309082, 0.040863037, 0.015838623, -0.053710938, 0.016220093, -0.05029297, -0.025772095, 0.019134521, 0.008628845, -0.013893127, 0.015686035, 0.014602661, -0.008026123, 0.010551453, -0.014404297, -0.010826111, 0.053619385, 0.0040779114, -0.019760132, -0.0209...
http://arxiv.org/abs/2006.09235v1
http://arxiv.org/pdf/2006.09235v1
Weakly-supervised Domain Adaption for Aspect Extraction via Multi-level Interaction Transfer
Tao Liang, Wenya Wang, Fengmao Lv
cs.CL
cs.CL
Fine-grained aspect extraction is an essential sub-task in aspect based opinion analysis. It aims to identify the aspect terms (a.k.a. opinion targets) of a product or service in each sentence. However, expensive annotation process is usually involved to acquire sufficient token-level labels for each domain. To address...
2020-06-16T15:11:51
2020-06-16T15:11:51
[ 0.024749756, -0.0138549805, -0.031097412, -0.060333252, -0.026245117, -0.0028038025, -0.048736572, 0.008300781, 0.029251099, 0.024917603, 0.03491211, -0.010948181, 0.04019165, -0.033966064, 0.0024700165, 0.005756378, -0.010154724, 0.027252197, -0.037231445, -0.036376953, -0.0...
[ 0.041290283, -0.04727173, 0.0026245117, -0.037750244, -0.019546509, -0.0045204163, -0.02217102, -0.004585266, 0.04611206, 0.025466919, 0.0005745888, -0.00013029575, 0.0335083, -0.0021877289, -0.013801575, -0.020980835, -0.011741638, 0.049072266, -0.010482788, -0.03717041, -0....
http://arxiv.org/abs/2006.09242v3
http://arxiv.org/pdf/2006.09242v3
Modeling Graph Structure via Relative Position for Text Generation from Knowledge Graphs
Martin Schmitt, Leonardo F. R. Ribeiro, Philipp Dufter, Iryna Gurevych, Hinrich Schütze
cs.CL
cs.CL
We present Graformer, a novel Transformer-based encoder-decoder architecture for graph-to-text generation. With our novel graph self-attention, the encoding of a node relies on all nodes in the input graph - not only direct neighbors - facilitating the detection of global patterns. We represent the relation between two...
2021-04-27T09:13:08
2020-06-16T15:20:04
[ -0.051849365, -0.043029785, -0.047912598, -0.028137207, -0.07867432, 0.023910522, -0.09692383, 0.013267517, 0.0043411255, 0.03555298, -0.017242432, -0.0046424866, 0.018432617, -0.007408142, -0.0184021, -0.013893127, -0.000872612, 0.018417358, -0.0027065277, -0.010093689, -0.0...
[ -0.014015198, -0.04272461, 0.005695343, -0.009391785, -0.06958008, 0.015464783, -0.07989502, -0.0056877136, 0.023651123, 0.034606934, -0.00680542, 0.013656616, 0.029846191, -0.016601562, 0.021636963, 0.0024089813, 0.0031204224, 0.04055786, 0.010810852, -0.021896362, -0.003152...
http://arxiv.org/abs/2006.09277v1
http://arxiv.org/pdf/2006.09277v1
Communicative need modulates competition in language change
Andres Karjus, Richard A. Blythe, Simon Kirby, Kenny Smith
cs.CL
cs.CL
All living languages change over time. The causes for this are many, one being the emergence and borrowing of new linguistic elements. Competition between the new elements and older ones with a similar semantic or grammatical function may lead to speakers preferring one of them, and leaving the other to go out of use. ...
2020-06-16T16:11:04
2020-06-16T16:11:04
[ 0.0042381287, -0.02407837, 0.00037121773, -0.009399414, -0.03262329, -0.0028705597, -0.02067566, 0.006752014, -0.052001953, 0.008270264, 0.0044670105, 0.0051841736, -0.017074585, -0.03427124, 0.035827637, -0.044647217, -0.015533447, 0.020309448, -0.006412506, -0.0418396, -0.0...
[ 0.047027588, -0.014839172, -0.016906738, 0.00932312, 0.014083862, -0.02885437, -0.022140503, -0.022201538, 0.0053215027, 0.020019531, -0.038879395, 0.027023315, 0.018600464, 0.0019893646, 0.02331543, -0.03729248, -0.010925293, 0.041168213, 0.010795593, -0.030685425, -0.011421...
http://arxiv.org/abs/2006.09336v2
http://arxiv.org/pdf/2006.09336v2
Cross-Cultural Similarity Features for Cross-Lingual Transfer Learning of Pragmatically Motivated Tasks
Jimin Sun, Hwijeen Ahn, Chan Young Park, Yulia Tsvetkov, David R. Mortensen
cs.CL
cs.CL
Much work in cross-lingual transfer learning explored how to select better transfer languages for multilingual tasks, primarily focusing on typological and genealogical similarities between languages. We hypothesize that these measures of linguistic proximity are not enough when working with pragmatically-motivated tas...
2021-04-08T08:31:54
2020-06-16T17:20:25
[ 0.01210022, -0.005493164, -0.008224487, 0.003730774, -0.015075684, -0.009727478, -0.01109314, -0.026016235, 0.006416321, 0.020721436, -0.0053977966, 0.016448975, 0.029083252, -0.027740479, 0.021453857, -0.0065231323, -0.007171631, 0.039245605, -0.0021648407, -0.02722168, -0.0...
[ 0.033935547, -0.018417358, 0.01473999, 0.017028809, -0.018936157, -0.003074646, -0.03756714, -0.043701172, 0.03866577, 0.034240723, -0.0030269623, 0.0059432983, 0.05697632, -0.0060691833, -0.006427765, 0.0064353943, 0.006061554, 0.030090332, 0.023635864, -0.014076233, -0.0405...
http://arxiv.org/abs/2006.09432v1
http://arxiv.org/pdf/2006.09432v1
The Role of Verb Semantics in Hungarian Verb-Object Order
Dorottya Demszky, László Kálmán, Dan Jurafsky, Beth Levin
cs.CL
cs.CL
Hungarian is often referred to as a discourse-configurational language, since the structural position of constituents is determined by their logical function (topic or comment) rather than their grammatical function (e.g., subject or object). We build on work by Koml\'osy (1989) and argue that in addition to discourse ...
2020-06-16T18:23:58
2020-06-16T18:23:58
[ 0.039276123, 0.005191803, -0.013931274, -0.072021484, -0.0059509277, -0.00522995, -0.036590576, 0.006061554, -0.009483337, 0.078430176, 0.014266968, -0.016571045, 0.08886719, -0.017288208, -0.028457642, 0.01739502, -0.013259888, 0.0345459, -0.016235352, -0.015151978, 0.010871...
[ 0.0357666, -0.0018119812, -0.008560181, -0.041870117, 0.00712204, 0.00027108192, -0.050445557, -0.018798828, -0.008674622, 0.057891846, -0.014373779, 0.010414124, 0.06591797, 0.0023860931, -0.013420105, -0.0046806335, -0.0256958, 0.044525146, -0.012832642, -0.039611816, 0.005...
http://arxiv.org/abs/2006.09589v2
http://arxiv.org/pdf/2006.09589v2
Modeling Subjective Assessments of Guilt in Newspaper Crime Narratives
Elisa Kreiss, Zijian Wang, Christopher Potts
cs.CL
cs.CL
Crime reporting is a prevalent form of journalism with the power to shape public perceptions and social policies. How does the language of these reports act on readers? We seek to address this question with the SuspectGuilt Corpus of annotated crime stories from English-language newspapers in the U.S. For SuspectGuilt,...
2020-10-14T22:38:44
2020-06-17T01:21:19
[ 0.0023975372, -0.030471802, 0.019958496, 0.019989014, -0.05050659, -0.0036468506, 0.040039062, -0.021835327, 0.00983429, -0.0043525696, -0.009346008, -0.007865906, -0.027954102, -0.027389526, -0.014381409, -0.027740479, -0.024917603, -0.0026435852, -0.044158936, -0.04650879, ...
[ 0.0031147003, -0.0519104, 0.012908936, 0.0152282715, -0.053527832, 0.00024104118, -0.005542755, -0.018249512, -0.00054216385, 0.023361206, 0.004901886, 0.0049476624, 0.009353638, -0.0060577393, 0.0051994324, -0.018295288, -0.024032593, 0.022918701, 0.00068616867, -0.045013428, ...
http://arxiv.org/abs/2006.09627v2
http://arxiv.org/pdf/2006.09627v2
Building Low-Resource NER Models Using Non-Speaker Annotation
Tatiana Tsygankova, Francesca Marini, Stephen Mayhew, Dan Roth
cs.CL
cs.CL
In low-resource natural language processing (NLP), the key problems are a lack of target language training data, and a lack of native speakers to create it. Cross-lingual methods have had notable success in addressing these concerns, but in certain common circumstances, such as insufficient pre-training corpora or lang...
2021-04-26T16:28:48
2020-06-17T03:24:38
[ 0.015853882, -0.032440186, 0.015914917, -0.04159546, -0.058563232, 0.026260376, -0.06829834, -0.004589081, 0.0051345825, 0.025512695, -0.002948761, -0.0008831024, 0.0021686554, -0.029083252, -0.007286072, -0.006511688, -0.013221741, 0.03616333, 0.00016093254, -0.020706177, 0....
[ 0.045654297, -0.023986816, 0.011962891, -0.020889282, -0.018310547, -0.0032444, -0.05230713, -0.018920898, 0.02784729, 0.021347046, -0.027816772, 0.0033035278, 0.01826477, 0.007850647, 0.0066833496, -0.018005371, -0.018707275, 0.06021118, 0.02330017, -0.024032593, -0.01554870...
http://arxiv.org/abs/2006.09639v1
http://arxiv.org/pdf/2006.09639v1
Iterative Edit-Based Unsupervised Sentence Simplification
Dhruv Kumar, Lili Mou, Lukasz Golab, Olga Vechtomova
cs.CL
cs.CL
We present a novel iterative, edit-based approach to unsupervised sentence simplification. Our model is guided by a scoring function involving fluency, simplicity, and meaning preservation. Then, we iteratively perform word and phrase-level edits on the complex sentence. Compared with previous approaches, our model doe...
2020-06-17T03:53:12
2020-06-17T03:53:12
[ -0.02810669, -0.0024585724, -0.028717041, -0.043060303, -0.025939941, 0.028961182, -0.09222412, -0.009536743, 0.029891968, 0.05102539, 0.01134491, -0.0014724731, 0.021438599, -0.011787415, -0.036499023, -0.0010843277, -0.0072746277, -0.0038795471, -0.05142212, -0.017196655, -...
[ -0.008735657, -0.0061683655, -0.0020275116, -0.026306152, -0.020767212, 0.022155762, -0.09429932, -0.023529053, 0.015029907, 0.041534424, 0.01398468, -0.012039185, 0.048919678, -0.010299683, -0.008987427, -0.000562191, -0.0055618286, 0.029022217, -0.015357971, -0.037628174, -...
http://arxiv.org/abs/2006.09685v1
http://arxiv.org/pdf/2006.09685v1
Exploiting Review Neighbors for Contextualized Helpfulness Prediction
Jiahua Du, Jia Rong, Hua Wang, Yanchun Zhang
cs.CL
cs.CL
Helpfulness prediction techniques have been widely used to identify and recommend high-quality online reviews to customers. Currently, the vast majority of studies assume that a review's helpfulness is self-contained. In practice, however, customers hardly process reviews independently given the sequential nature. The ...
2020-06-17T07:02:42
2020-06-17T07:02:42
[ 0.04586792, -0.011489868, 0.01940918, -0.049194336, -0.036102295, 0.00060653687, -0.045166016, -0.0284729, 0.0048599243, 0.045837402, -0.04525757, 0.03173828, 0.035980225, 0.009552002, 0.0049552917, -0.0064811707, 0.0023994446, 0.027816772, 0.0041770935, -0.024017334, -0.0153...
[ 0.061706543, -0.031143188, 0.040039062, -0.010948181, -0.03186035, -0.015914917, -0.0017671585, -0.036956787, -0.007881165, 0.048431396, -0.036590576, 0.027664185, 0.03201294, 0.02468872, 0.014228821, -0.020751953, 0.005924225, 0.04232788, 0.029327393, -0.028549194, -0.014190...
http://arxiv.org/abs/2006.09719v1
http://arxiv.org/pdf/2006.09719v1
Automatically Ranked Russian Paraphrase Corpus for Text Generation
Vadim Gudkov, Olga Mitrofanova, Elizaveta Filippskikh
cs.CL
cs.CL
The article is focused on automatic development and ranking of a large corpus for Russian paraphrase generation which proves to be the first corpus of such type in Russian computational linguistics. Existing manually annotated paraphrase datasets for Russian are limited to small-sized ParaPhraser corpus and ParaPlag wh...
2020-06-17T08:40:52
2020-06-17T08:40:52
[ 0.009284973, -0.04119873, 0.030151367, -0.035614014, -0.041015625, 0.041625977, -0.047210693, -0.014472961, 0.017562866, 0.07556152, -0.0149002075, -0.0029792786, 0.05581665, -0.0079422, -0.02947998, -0.004032135, -0.032592773, 0.05001831, -0.043548584, -0.021072388, -0.02061...
[ 0.014770508, -0.030899048, 0.018096924, -0.011558533, -0.018096924, 0.01133728, -0.041534424, -0.022872925, 0.014350891, 0.035247803, -0.027954102, 0.0209198, 0.051483154, 0.005519867, 0.008773804, -0.017822266, -0.037322998, 0.06359863, -0.021057129, -0.0256958, -0.001142501...
http://arxiv.org/abs/2006.09766v1
http://arxiv.org/pdf/2006.09766v1
Improving unsupervised neural aspect extraction for online discussions using out-of-domain classification
Anton Alekseev, Elena Tutubalina, Valentin Malykh, Sergey Nikolenko
cs.CL
cs.CL
Deep learning architectures based on self-attention have recently achieved and surpassed state of the art results in the task of unsupervised aspect extraction and topic modeling. While models such as neural attention-based aspect extraction (ABAE) have been successfully applied to user-generated texts, they are less c...
2020-06-17T10:34:16
2020-06-17T10:34:16
[ 0.055786133, -0.0038719177, 0.0040740967, -0.024978638, -0.011276245, -0.005680084, -0.0035438538, -0.01751709, 0.04058838, 0.022918701, 0.034729004, -0.01247406, 0.0053520203, -0.027267456, -0.015525818, -0.015701294, -0.020706177, 0.03137207, -0.02027893, -0.02545166, -0.02...
[ 0.04638672, -0.007965088, -0.0042381287, -0.016708374, -0.024398804, -0.01713562, -0.02708435, -0.030349731, 0.05517578, 0.062286377, 0.01448822, 0.008605957, 0.04321289, 0.0010280609, -0.03491211, 0.014045715, -0.02243042, 0.018066406, -0.023498535, -0.036102295, -0.01770019...
http://arxiv.org/abs/2006.10093v1
http://arxiv.org/pdf/2006.10093v1
Extensively Matching for Few-shot Learning Event Detection
Viet Dac Lai, Franck Dernoncourt, Thien Huu Nguyen
cs.CL
cs.CL
Current event detection models under super-vised learning settings fail to transfer to newevent types. Few-shot learning has not beenexplored in event detection even though it al-lows a model to perform well with high gener-alization on new event types. In this work, weformulate event detection as a few-shot learn-ing ...
2020-06-17T18:30:30
2020-06-17T18:30:30
[ 0.03967285, 0.0030593872, 0.0044517517, -0.0034561157, -0.006729126, -0.019897461, 0.0057868958, -0.04925537, 0.006641388, 0.0002875328, -0.031707764, 0.025161743, -0.02444458, -0.023040771, 0.020080566, -0.0340271, 0.012741089, 0.016662598, -0.016082764, -0.022567749, 0.0135...
[ 0.06390381, -0.021499634, 0.005760193, -0.0036525726, -0.010673523, -0.028396606, -0.011512756, -0.042785645, 0.025878906, 0.012252808, -0.01751709, 0.016235352, 0.0063323975, -0.012252808, 0.004676819, -0.02281189, 0.016220093, -0.009239197, 0.00073719025, -0.028198242, 0.04...
http://arxiv.org/abs/2006.10157v1
http://arxiv.org/pdf/2006.10157v1
Is this Dialogue Coherent? Learning from Dialogue Acts and Entities
Alessandra Cervone, Giuseppe Riccardi
cs.CL
cs.CL
In this work, we investigate the human perception of coherence in open-domain dialogues. In particular, we address the problem of annotating and modeling the coherence of next-turn candidates while considering the entire history of the dialogue. First, we create the Switchboard Coherence (SWBD-Coh) corpus, a dataset of...
2020-06-17T21:02:40
2020-06-17T21:02:40
[ -0.00060367584, 0.0023555756, -0.051635742, -0.039398193, -0.0008878708, -0.015144348, -0.019165039, 0.016036987, 0.012428284, 0.009407043, -0.010116577, -0.020401001, -0.015068054, -0.033203125, -0.03930664, -0.036468506, -0.008583069, 0.061431885, 0.01574707, -0.01612854, 0...
[ 0.05090332, 0.0058517456, -0.023605347, -0.039398193, 0.010421753, -0.028366089, -0.04852295, -0.026947021, 0.004753113, 0.060180664, -0.0034942627, -0.0025901794, 0.07080078, 0.0031585693, -0.034057617, -0.016342163, -0.04232788, 0.035888672, -0.019500732, -0.033233643, 0.00...
http://arxiv.org/abs/2006.10270v2
http://arxiv.org/pdf/2006.10270v2
Multi-branch Attentive Transformer
Yang Fan, Shufang Xie, Yingce Xia, Lijun Wu, Tao Qin, Xiang-Yang Li, Tie-Yan Liu
cs.CL
cs.CL
While the multi-branch architecture is one of the key ingredients to the success of computer vision tasks, it has not been well investigated in natural language processing, especially sequence learning tasks. In this work, we propose a simple yet effective variant of Transformer called multi-branch attentive Transforme...
2020-07-26T13:04:49
2020-06-18T04:24:28
[ 0.033477783, -0.0046195984, 0.025177002, -0.019134521, -0.017486572, 0.011924744, -0.0035037994, -0.017868042, 0.01878357, 0.050872803, 0.013923645, 0.011016846, 0.017669678, -0.023864746, -0.010643005, 0.0071258545, 0.025177002, 0.03262329, 0.007648468, -0.0011558533, -0.002...
[ 0.031433105, 0.004436493, 0.0023479462, -0.025390625, -0.014038086, -0.0033054352, -0.048675537, -0.0131073, 0.03314209, 0.036102295, -0.035461426, -0.012779236, 0.04776001, 0.0059013367, -0.015357971, -0.013641357, 0.0134887695, 0.04119873, 0.010665894, -0.021057129, -0.0070...