id
stringlengths
33
34
title
stringlengths
6
218
authors
stringlengths
5
6.98k
categories
stringlengths
5
188
summary
stringlengths
34
3.26k
link_pdf
stringlengths
33
34
updated
timestamp[s]
published
timestamp[s]
embeddings
dict
http://arxiv.org/abs/2301.13298v1
LongEval: Guidelines for Human Evaluation of Faithfulness in Long-form Summarization
Kalpesh Krishna, Erin Bransom, Bailey Kuehl, Mohit Iyyer, Pradeep Dasigi, Arman Cohan, Kyle Lo
cs.CL
While human evaluation remains best practice for accurately judging the faithfulness of automatically-generated summaries, few solutions exist to address the increased difficulty and workload when evaluating long-form summaries. Through a survey of 162 papers on long-form summarization, we first shed light on current h...
http://arxiv.org/pdf/2301.13298v1
2023-01-30T21:31:48
2023-01-30T21:31:48
{ "title": [ -0.0049743652, 0.016571045, -0.04067993, -0.067871094, 0.0025596619, -0.010482788, -0.042236328, -0.028045654, -0.011238098, 0.048553467, -0.016098022, 0.013305664, 0.0126953125, -0.0075531006, -0.047729492, 0.006587982, -0.00015234947, ...
http://arxiv.org/abs/2301.13294v3
Adaptive Machine Translation with Large Language Models
Yasmin Moslem, Rejwanul Haque, John D. Kelleher, Andy Way
cs.CL
Consistency is a key requirement of high-quality translation. It is especially important to adhere to pre-approved terminology and adapt to corrected translations in domain-specific projects. Machine translation (MT) has achieved significant progress in the area of domain adaptation. However, real-time adaptation remai...
http://arxiv.org/pdf/2301.13294v3
2023-05-09T11:04:43
2023-01-30T21:17:15
{ "title": [ -0.041503906, -0.012969971, -0.021835327, -0.05001831, -0.0012607574, 0.016784668, -0.04534912, 0.006603241, 0.0065078735, 0.045135498, -0.0056991577, -0.01007843, 0.025863647, -0.018676758, -0.04434204, 0.025131226, 0.0022125244, 0....
http://arxiv.org/abs/2301.13268v2
Contextual Dynamic Prompting for Response Generation in Task-oriented Dialog Systems
Sandesh Swamy, Narges Tabari, Chacha Chen, Rashmi Gangadharaiah
cs.CL
Response generation is one of the critical components in task-oriented dialog systems. Existing studies have shown that large pre-trained language models can be adapted to this task. The typical paradigm of adapting such extremely large language models would be by fine-tuning on the downstream tasks which is not only t...
http://arxiv.org/pdf/2301.13268v2
2023-02-10T23:16:02
2023-01-30T20:26:02
{ "title": [ -0.0056915283, -0.007881165, -0.037506104, -0.082092285, 0.007827759, 0.0017976761, -0.04724121, 0.012107849, -0.0046195984, 0.03567505, -0.004463196, 0.0149383545, -0.01373291, -0.016067505, -0.05618286, -0.003955841, 0.013534546, -...
http://arxiv.org/abs/2301.13267v1
ArchiSound: Audio Generation with Diffusion
Flavio Schneider
cs.SD, cs.CL, eess.AS
The recent surge in popularity of diffusion models for image generation has brought new attention to the potential of these models in other areas of media generation. One area that has yet to be fully explored is the application of diffusion models to audio generation. Audio generation requires an understanding of mult...
http://arxiv.org/pdf/2301.13267v1
2023-01-30T20:23:26
2023-01-30T20:23:26
{ "title": [ -0.03137207, -0.015335083, -0.048919678, -0.046539307, 0.0012292862, -0.0063476562, -0.026489258, -0.007774353, 0.008323669, 0.03982544, 0.0022201538, 0.010253906, 0.016296387, -0.03475952, -0.034423828, 0.022064209, -0.011985779, 0....
http://arxiv.org/abs/2301.13155v2
Advancing Radiograph Representation Learning with Masked Record Modeling
Hong-Yu Zhou, Chenyu Lian, Liansheng Wang, Yizhou Yu
cs.CV, cs.CL, cs.LG
Modern studies in radiograph representation learning rely on either self-supervision to encode invariant semantics or associated radiology reports to incorporate medical expertise, while the complementarity between them is barely noticed. To explore this, we formulate the self- and report-completion as two complementar...
http://arxiv.org/pdf/2301.13155v2
2023-02-15T07:33:35
2023-01-30T18:33:32
{ "title": [ -0.02986145, 0.004436493, -0.06549072, -0.05886841, -0.011016846, -0.0025138855, -0.06750488, 0.015304565, -0.005088806, 0.0049057007, -0.036071777, -0.015975952, -0.030349731, -0.04425049, -0.01600647, -0.010681152, 0.00774765, 0.04...
http://arxiv.org/abs/2301.13154v2
Protein Representation Learning via Knowledge Enhanced Primary Structure Modeling
Hong-Yu Zhou, Yunxiang Fu, Zhicheng Zhang, Cheng Bian, Yizhou Yu
cs.LG, cs.CL, cs.MM
Protein representation learning has primarily benefited from the remarkable development of language models (LMs). Accordingly, pre-trained protein models also suffer from a problem in LMs: a lack of factual knowledge. The recent solution models the relationships between protein and associated knowledge terms as the kno...
http://arxiv.org/pdf/2301.13154v2
2023-02-15T07:26:47
2023-01-30T18:33:15
{ "title": [ -0.028121948, -0.014480591, -0.04550171, -0.07098389, -0.01361084, -0.0005402565, -0.03845215, 0.02645874, 0.026519775, 0.019348145, -0.0049705505, -0.0049819946, -0.005306244, -0.033996582, -0.028503418, 0.017547607, 0.015266418, 0....
http://arxiv.org/abs/2301.13126v2
LEXTREME: A Multi-Lingual and Multi-Task Benchmark for the Legal Domain
Joel Niklaus, Veton Matoshi, Pooja Rani, Andrea Galassi, Matthias Stürmer, Ilias Chalkidis
cs.CL, cs.AI, cs.LG, 68T50, I.2
Lately, propelled by the phenomenal advances around the transformer architecture, the legal NLP field has enjoyed spectacular growth. To measure progress, well curated and challenging benchmarks are crucial. However, most benchmarks are English only and in legal NLP specifically there is no multilingual benchmark avail...
http://arxiv.org/pdf/2301.13126v2
2023-10-31T23:05:31
2023-01-30T18:05:08
{ "title": [ -0.044189453, -0.015808105, -0.05697632, -0.059173584, 0.003376007, 0.0020599365, -0.06188965, 0.0018434525, 0.012123108, 0.023956299, -0.006717682, 0.012207031, -0.004486084, -0.016586304, -0.037200928, 0.010047913, -0.0184021, 0.04...
http://arxiv.org/abs/2301.13089v1
Can an AI Win Ghana's National Science and Maths Quiz? An AI Grand Challenge for Education
George Boateng, Victor Kumbol, Elsie Effah Kaufmann
cs.CL, cs.AI, cs.HC
There is a lack of enough qualified teachers across Africa which hampers efforts to provide adequate learning support such as educational question answering (EQA) to students. An AI system that can enable students to ask questions via text or voice and get instant answers will make high-quality education accessible. De...
http://arxiv.org/pdf/2301.13089v1
2023-01-30T17:28:33
2023-01-30T17:28:33
{ "title": [ -0.00680542, -0.022384644, 0.0039482117, 0.0072288513, -0.018630981, -0.011947632, 0.044311523, -0.01512146, -0.034179688, 0.035339355, -0.044921875, 0.034729004, 0.0146102905, -0.018157959, -0.01197052, -0.023101807, 0.0008454323, -...
http://arxiv.org/abs/2301.13083v2
Communication Drives the Emergence of Language Universals in Neural Agents: Evidence from the Word-order/Case-marking Trade-off
Yuchen Lian, Arianna Bisazza, Tessa Verhoef
cs.CL, cs.AI
Artificial learners often behave differently from human learners in the context of neural agent-based simulations of language emergence and change. A common explanation is the lack of appropriate cognitive biases in these learners. However, it has also been proposed that more naturalistic settings of language learning ...
http://arxiv.org/pdf/2301.13083v2
2023-06-01T03:54:09
2023-01-30T17:22:33
{ "title": [ -0.010665894, 0.010475159, -0.03692627, -0.017105103, 0.0025901794, -0.018081665, -0.016677856, -0.0039024353, -0.02558899, 0.0423584, 0.010803223, 0.005039215, 0.023345947, -0.03189087, -0.023635864, -0.003929138, -0.03048706, 0.054...
http://arxiv.org/abs/2301.13066v2
A Human Word Association based model for topic detection in social networks
Mehrdad Ranjbar Khadivi, Shahin Akbarpour, Mohammad-Reza Feizi-Derakhshi, Babak Anari
cs.CL, cs.AI
With the widespread use of social networks, detecting the topics discussed in these networks has become a significant challenge. The current works are mainly based on frequent pattern mining or semantic relations, and the language structure is not considered. The meaning of language structural methods is to discover th...
http://arxiv.org/pdf/2301.13066v2
2023-07-18T11:39:22
2023-01-30T17:10:34
{ "title": [ -0.017669678, 0.019973755, -0.040924072, -0.012016296, 0.011741638, -0.010643005, -0.023666382, -0.044067383, 0.01197052, 0.03225708, -0.021621704, 0.028900146, 0.0006170273, -0.0005340576, -0.028701782, 0.0079422, -0.022613525, 0.03...
http://arxiv.org/abs/2301.13042v1
The Secret of Metaphor on Expressing Stronger Emotion
Yucheng Li, Frank Guerin, Chenghua Lin
cs.CL
Metaphors are proven to have stronger emotional impact than literal expressions. Although this conclusion is shown to be promising in benefiting various NLP applications, the reasons behind this phenomenon are not well studied. This paper conducts the first study in exploring how metaphors convey stronger emotion than ...
http://arxiv.org/pdf/2301.13042v1
2023-01-30T16:36:02
2023-01-30T16:36:02
{ "title": [ 0.025299072, -0.005607605, 0.0048561096, 0.013954163, -0.02255249, -0.011192322, 0.03778076, -0.032684326, -0.039154053, 0.04928589, -0.012077332, -0.0052986145, 0.04272461, -0.044525146, 0.007980347, -0.008331299, 0.018554688, 0.027...
http://arxiv.org/abs/2301.13039v1
Representation biases in sentence transformers
Dmitry Nikolaev, Sebastian Padó
cs.CL
Variants of the BERT architecture specialised for producing full-sentence representations often achieve better performance on downstream tasks than sentence embeddings extracted from vanilla BERT. However, there is still little understanding of what properties of inputs determine the properties of such representations....
http://arxiv.org/pdf/2301.13039v1
2023-01-30T16:35:23
2023-01-30T16:35:23
{ "title": [ 0.016616821, 0.002336502, -0.010391235, -0.042022705, -0.0028133392, -0.011451721, -0.015792847, -0.023803711, 0.0011281967, 0.042663574, -0.045837402, 0.0012969971, 0.022384644, -0.0010595322, -0.044799805, -0.006580353, -0.025314331, ...
http://arxiv.org/abs/2301.13003v2
Knowledge Transfer from Pre-trained Language Models to Cif-based Speech Recognizers via Hierarchical Distillation
Minglun Han, Feilong Chen, Jing Shi, Shuang Xu, Bo Xu
cs.CL, cs.AI, cs.SD, eess.AS
Large-scale pre-trained language models (PLMs) have shown great potential in natural language processing tasks. Leveraging the capabilities of PLMs to enhance automatic speech recognition (ASR) systems has also emerged as a promising research direction. However, previous works may be limited by the inflexible structure...
http://arxiv.org/pdf/2301.13003v2
2023-05-28T16:13:52
2023-01-30T15:44:55
{ "title": [ -0.008987427, -0.0090408325, -0.047607422, -0.0960083, -0.008460999, -0.0008921623, -0.074645996, 0.026382446, 0.009521484, 0.025848389, -0.032714844, 0.004837036, -0.006134033, -0.012382507, -0.04727173, -0.018661499, -0.0073394775, ...
http://arxiv.org/abs/2301.12971v2
Quantifying Context Mixing in Transformers
Hosein Mohebbi, Willem Zuidema, Grzegorz Chrupała, Afra Alishahi
cs.CL, cs.LG
Self-attention weights and their transformed variants have been the main source of information for analyzing token-to-token interactions in Transformer-based models. But despite their ease of interpretation, these weights are not faithful to the models' decisions as they are only one part of an encoder, and other compo...
http://arxiv.org/pdf/2301.12971v2
2023-02-08T10:46:38
2023-01-30T15:19:02
{ "title": [ 0.01259613, 0.022445679, -0.046875, -0.06732178, 0.037841797, -0.01373291, -0.010025024, -0.018722534, -0.023635864, 0.070495605, -0.021484375, 0.023773193, 0.024887085, -0.0151901245, -0.021560669, 0.010070801, 0.0051574707, 0.03964...
http://arxiv.org/abs/2301.12969v1
Using n-aksaras to model Sanskrit and Sanskrit-adjacent texts
Charles Li
cs.CL
Despite -- or perhaps because of -- their simplicity, n-grams, or contiguous sequences of tokens, have been used with great success in computational linguistics since their introduction in the late 20th century. Recast as k-mers, or contiguous sequences of monomers, they have also found applications in computational bi...
http://arxiv.org/pdf/2301.12969v1
2023-01-30T15:17:06
2023-01-30T15:17:06
{ "title": [ 0.00031661987, 0.019546509, -0.017593384, -0.04232788, -0.023666382, 0.0029468536, -0.049468994, -0.02104187, 0.0013685226, 0.033233643, -0.047790527, 0.024124146, 0.005290985, -0.007457733, -0.022857666, -0.015533447, 0.008636475, 0...
http://arxiv.org/abs/2301.12920v4
Active Learning for Multilingual Semantic Parser
Zhuang Li, Gholamreza Haffari
cs.CL
Current multilingual semantic parsing (MSP) datasets are almost all collected by translating the utterances in the existing datasets from the resource-rich language to the target language. However, manual translation is costly. To reduce the translation effort, this paper proposes the first active learning procedure fo...
http://arxiv.org/pdf/2301.12920v4
2023-10-11T04:08:26
2023-01-30T14:19:29
{ "title": [ -0.023544312, 0.00497818, -0.019073486, -0.080566406, 0.01247406, 0.018676758, -0.064453125, 0.004802704, -0.0046806335, 0.042633057, 0.0019140244, 0.008605957, -0.01928711, -0.018508911, -0.03125, 0.005367279, -0.005039215, 0.047637...
http://arxiv.org/abs/2301.12868v3
On Robustness of Prompt-based Semantic Parsing with Large Pre-trained Language Model: An Empirical Study on Codex
Terry Yue Zhuo, Zhuang Li, Yujin Huang, Fatemeh Shiri, Weiqing Wang, Gholamreza Haffari, Yuan-Fang Li
cs.CL
Semantic parsing is a technique aimed at constructing a structured representation of the meaning of a natural-language question. Recent advancements in few-shot language models trained on code have demonstrated superior performance in generating these representations compared to traditional unimodal language models, wh...
http://arxiv.org/pdf/2301.12868v3
2023-03-09T11:01:02
2023-01-30T13:21:00
{ "title": [ -0.0082092285, -0.0019435883, -0.026916504, -0.05984497, -0.010498047, -0.0044822693, -0.078430176, -0.016342163, -0.011276245, 0.013244629, -0.038238525, 0.009376526, -0.009155273, -0.009803772, -0.03643799, -0.019180298, -0.0034332275,...
http://arxiv.org/abs/2301.12867v4
Red teaming ChatGPT via Jailbreaking: Bias, Robustness, Reliability and Toxicity
Terry Yue Zhuo, Yujin Huang, Chunyang Chen, Zhenchang Xing
cs.CL, cs.SE
Recent breakthroughs in natural language processing (NLP) have permitted the synthesis and comprehension of coherent text in an open-ended way, therefore translating the theoretical algorithms into practical applications. The large language models (LLMs) have significantly impacted businesses such as report summarizati...
http://arxiv.org/pdf/2301.12867v4
2023-05-29T17:46:54
2023-01-30T13:20:48
{ "title": [ 0.0042533875, 0.0015974045, -0.0078125, 0.008766174, -0.014213562, -0.008422852, -0.0016584396, -0.032226562, -0.05557251, 0.02482605, 0.0003015995, 0.01234436, -0.032104492, -0.040100098, -0.011917114, -0.024932861, -0.023132324, 0....
http://arxiv.org/abs/2301.12866v2
N-Gram Nearest Neighbor Machine Translation
Rui Lv, Junliang Guo, Rui Wang, Xu Tan, Qi Liu, Tao Qin
cs.CL, cs.LG
Nearest neighbor machine translation augments the Autoregressive Translation~(AT) with $k$-nearest-neighbor retrieval, by comparing the similarity between the token-level context representations of the target tokens in the query and the datastore. However, the token-level representation may introduce noise when transla...
http://arxiv.org/pdf/2301.12866v2
2023-02-07T14:15:26
2023-01-30T13:19:19
{ "title": [ -0.045898438, -0.02571106, 0.0066604614, -0.045288086, -0.017791748, 0.02355957, -0.043273926, -0.0098724365, 0.0010576248, 0.052581787, -0.017807007, 0.004135132, 0.03161621, -0.016830444, -0.037750244, 0.015510559, 0.005443573, 0.0...
http://arxiv.org/abs/2301.12855v1
How Far Can It Go?: On Intrinsic Gender Bias Mitigation for Text Classification
Ewoenam Tokpo, Pieter Delobelle, Bettina Berendt, Toon Calders
cs.CL
To mitigate gender bias in contextualized language models, different intrinsic mitigation strategies have been proposed, alongside many bias metrics. Considering that the end use of these language models is for downstream tasks like text classification, it is important to understand how these intrinsic bias mitigation ...
http://arxiv.org/pdf/2301.12855v1
2023-01-30T13:05:48
2023-01-30T13:05:48
{ "title": [ -0.022277832, 0.0038223267, -0.05657959, -0.032165527, 0.015731812, -0.028076172, -0.010360718, -0.008399963, 0.0067710876, 0.04135132, -0.03475952, 0.015411377, 0.0082473755, 0.012893677, -0.06210327, -0.019363403, -0.034118652, 0.0...
http://arxiv.org/abs/2301.12850v1
GE-Blender: Graph-Based Knowledge Enhancement for Blender
Xiaolei Lian, Xunzhu Tang, Yue Wang
cs.CL, cs.AI
Although the great success of open-domain dialogue generation, unseen entities can have a large impact on the dialogue generation task. It leads to performance degradation of the model in the dialog generation. Previous researches used retrieved knowledge of seen entities as the auxiliary data to enhance the representa...
http://arxiv.org/pdf/2301.12850v1
2023-01-30T13:00:20
2023-01-30T13:00:20
{ "title": [ -0.039794922, -0.0044174194, -0.07476807, -0.06124878, 0.008049011, -0.0023670197, -0.052703857, 0.011177063, 0.0006661415, 0.015274048, 0.010856628, 0.005970001, 0.005268097, -0.045959473, -0.010635376, 0.013412476, 0.0002872944, 0....
http://arxiv.org/abs/2301.12847v1
Finding the Law: Enhancing Statutory Article Retrieval via Graph Neural Networks
Antoine Louis, Gijs van Dijck, Gerasimos Spanakis
cs.IR, cs.CL
Statutory article retrieval (SAR), the task of retrieving statute law articles relevant to a legal question, is a promising application of legal text processing. In particular, high-quality SAR systems can improve the work efficiency of legal professionals and provide basic legal assistance to citizens in need at no co...
http://arxiv.org/pdf/2301.12847v1
2023-01-30T12:59:09
2023-01-30T12:59:09
{ "title": [ -0.0259552, -0.009063721, -0.038269043, -0.019805908, -0.01209259, -0.011489868, -0.021194458, -0.008987427, 0.023773193, 0.020462036, -0.052459717, 0.012573242, -0.016052246, 0.007385254, -0.059143066, -0.008773804, -0.0141067505, 0...
http://arxiv.org/abs/2301.12810v1
Crawling the Internal Knowledge-Base of Language Models
Roi Cohen, Mor Geva, Jonathan Berant, Amir Globerson
cs.CL, cs.AI
Language models are trained on large volumes of text, and as a result their parameters might contain a significant body of factual knowledge. Any downstream task performed by these models implicitly builds on these facts, and thus it is highly desirable to have means for representing this body of knowledge in an interp...
http://arxiv.org/pdf/2301.12810v1
2023-01-30T12:03:36
2023-01-30T12:03:36
{ "title": [ -0.012390137, -0.028564453, -0.015411377, -0.07714844, -0.027938843, 0.014320374, -0.06616211, 0.016479492, 0.0055274963, 0.038116455, -0.043518066, 0.0034313202, 0.007549286, -0.0046806335, -0.05029297, 0.005180359, 0.018432617, 0.0...
http://arxiv.org/abs/2301.12729v1
Response-act Guided Reinforced Dialogue Generation for Mental Health Counseling
Aseem Srivastava, Ishan Pandey, Md. Shad Akhtar, Tanmoy Chakraborty
cs.CL
Virtual Mental Health Assistants (VMHAs) have become a prevalent method for receiving mental health counseling in the digital healthcare space. An assistive counseling conversation commences with natural open-ended topics to familiarize the client with the environment and later converges into more fine-grained domain-s...
http://arxiv.org/pdf/2301.12729v1
2023-01-30T08:53:35
2023-01-30T08:53:35
{ "title": [ 0.006679535, -0.017349243, -0.031555176, -0.07092285, 0.0021152496, 0.006778717, -0.037384033, 0.022033691, -0.0048675537, 0.026138306, 0.016647339, 0.00819397, -0.019500732, -0.04812622, -0.032928467, -0.0085372925, -0.016845703, 0....
http://arxiv.org/abs/2301.12726v1
Specializing Smaller Language Models towards Multi-Step Reasoning
Yao Fu, Hao Peng, Litu Ou, Ashish Sabharwal, Tushar Khot
cs.CL, cs.AI, cs.LG
The surprising ability of Large Language Models (LLMs) to perform well on complex reasoning with only few-shot chain-of-thought prompts is believed to emerge only in very large-scale models (100+ billion parameters). We show that such abilities can, in fact, be distilled down from GPT-3.5 ($\ge$ 175B) to T5 variants ($...
http://arxiv.org/pdf/2301.12726v1
2023-01-30T08:51:19
2023-01-30T08:51:19
{ "title": [ -0.044647217, -0.035705566, -0.054901123, -0.0993042, -0.03277588, 0.021621704, -0.09857178, 0.035705566, -0.008460999, 0.028930664, -0.006187439, -0.0018157959, -0.001077652, -0.027694702, -0.02394104, 0.007789612, 0.007434845, 0.00...
http://arxiv.org/abs/2301.13852v1
ChatGPT or Human? Detect and Explain. Explaining Decisions of Machine Learning Model for Detecting Short ChatGPT-generated Text
Sandra Mitrović, Davide Andreoletti, Omran Ayoub
cs.CL, cs.LG
ChatGPT has the ability to generate grammatically flawless and seemingly-human replies to different types of questions from various domains. The number of its users and of its applications is growing at an unprecedented rate. Unfortunately, use and abuse come hand in hand. In this paper, we study whether a machine lear...
http://arxiv.org/pdf/2301.13852v1
2023-01-30T08:06:08
2023-01-30T08:06:08
{ "title": [ -0.043884277, 0.00094795227, 0.0034255981, 0.050323486, -0.068359375, -0.008056641, -0.006668091, -0.0362854, -0.06225586, 0.015167236, -0.03866577, 0.016784668, -0.012016296, -0.012756348, 0.03555298, -0.0435791, -0.036834717, 0.058...
http://arxiv.org/abs/2301.12715v1
Fine-Tuning Deteriorates General Textual Out-of-Distribution Detection by Distorting Task-Agnostic Features
Sishuo Chen, Wenkai Yang, Xiaohan Bi, Xu Sun
cs.CL
Detecting out-of-distribution (OOD) inputs is crucial for the safe deployment of natural language processing (NLP) models. Though existing methods, especially those based on the statistics in the feature space of fine-tuned pre-trained language models (PLMs), are claimed to be effective, their effectiveness on differen...
http://arxiv.org/pdf/2301.12715v1
2023-01-30T08:01:13
2023-01-30T08:01:13
{ "title": [ 0.020095825, -0.007537842, 0.00041556358, -0.013458252, -0.012123108, -0.017242432, 0.019515991, -0.010299683, -0.008407593, 0.041992188, -0.03753662, 0.0078125, 0.0020122528, -0.0007095337, -0.051605225, -0.030929565, -0.012504578, ...
http://arxiv.org/abs/2301.12711v2
UzbekTagger: The rule-based POS tagger for Uzbek language
Maksud Sharipov, Elmurod Kuriyozov, Ollabergan Yuldashev, Ogabek Sobirov
cs.CL
This research paper presents a part-of-speech (POS) annotated dataset and tagger tool for the low-resource Uzbek language. The dataset includes 12 tags, which were used to develop a rule-based POS-tagger tool. The corpus text used in the annotation process was made sure to be balanced over 20 different fields in order ...
http://arxiv.org/pdf/2301.12711v2
2023-03-01T14:31:12
2023-01-30T07:40:45
{ "title": [ 0.044647217, 0.016845703, -0.022384644, -0.030914307, 0.007965088, -0.021850586, -0.008361816, -0.02104187, -0.031433105, 0.060455322, -0.01272583, -0.01448822, 0.094177246, -0.042419434, 0.015029907, -0.01033783, 0.0006365776, -0.00...
http://arxiv.org/abs/2301.12699v1
KG-BERTScore: Incorporating Knowledge Graph into BERTScore for Reference-Free Machine Translation Evaluation
Zhanglin Wu, Min Zhang, Ming Zhu, Yinglu Li, Ting Zhu, Hao Yang, Song Peng, Ying Qin
cs.CL, cs.AI
BERTScore is an effective and robust automatic metric for referencebased machine translation evaluation. In this paper, we incorporate multilingual knowledge graph into BERTScore and propose a metric named KG-BERTScore, which linearly combines the results of BERTScore and bilingual named entity matching for reference-f...
http://arxiv.org/pdf/2301.12699v1
2023-01-30T07:12:15
2023-01-30T07:12:15
{ "title": [ -0.039215088, -0.031341553, -0.004714966, -0.05444336, -0.039978027, 0.021621704, -0.045928955, 0.012969971, -0.007926941, 0.049835205, -0.016204834, 0.0037612915, 0.032714844, -0.013069153, -0.060699463, -0.0008382797, -0.01222229, ...
http://arxiv.org/abs/2301.12660v2
Zero-shot Clarifying Question Generation for Conversational Search
Zhenduo Wang, Yuancheng Tu, Corby Rosset, Nick Craswell, Ming Wu, Qingyao Ai
cs.IR, cs.CL
A long-standing challenge for search and conversational assistants is query intention detection in ambiguous queries. Asking clarifying questions in conversational search has been widely studied and considered an effective solution to resolve query ambiguity. Existing work have explored various approaches for clarifyin...
http://arxiv.org/pdf/2301.12660v2
2023-02-10T17:05:21
2023-01-30T04:43:02
{ "title": [ 0.009727478, -0.015853882, -0.014320374, -0.042144775, -0.020889282, 0.0013093948, -0.03387451, -0.006580353, 0.03842163, 0.017364502, -0.012817383, -0.0022144318, -0.008460999, -0.020721436, -0.06188965, 0.0017766953, -0.017913818, ...
http://arxiv.org/abs/2301.12652v4
REPLUG: Retrieval-Augmented Black-Box Language Models
Weijia Shi, Sewon Min, Michihiro Yasunaga, Minjoon Seo, Rich James, Mike Lewis, Luke Zettlemoyer, Wen-tau Yih
cs.CL
We introduce REPLUG, a retrieval-augmented language modeling framework that treats the language model (LM) as a black box and augments it with a tuneable retrieval model. Unlike prior retrieval-augmented LMs that train language models with special cross attention mechanisms to encode the retrieved text, REPLUG simply p...
http://arxiv.org/pdf/2301.12652v4
2023-05-24T05:08:07
2023-01-30T04:18:09
{ "title": [ -0.019454956, -0.019454956, -0.034088135, -0.0647583, -0.008514404, 0.007843018, -0.06185913, -0.0035591125, 0.019454956, 0.032806396, -0.023269653, 0.0128479, 0.0031204224, 0.0026130676, -0.05822754, 0.0105896, -0.0146102905, 0.0414...
http://arxiv.org/abs/2301.12642v1
Exploring the Constructicon: Linguistic Analysis of a Computational CxG
Jonathan Dunn
cs.CL
Recent work has formulated the task for computational construction grammar as producing a constructicon given a corpus of usage. Previous work has evaluated these unsupervised grammars using both internal metrics (for example, Minimum Description Length) and external metrics (for example, performance on a dialectology ...
http://arxiv.org/pdf/2301.12642v1
2023-01-30T03:51:08
2023-01-30T03:51:08
{ "title": [ -0.044769287, -0.028121948, -0.035614014, -0.0057258606, -0.04510498, -0.010223389, -0.05154419, -0.030426025, -0.019500732, 0.036865234, -0.037963867, 0.017227173, -0.014091492, -0.020751953, -0.024093628, -0.020904541, -0.010826111, ...
http://arxiv.org/abs/2301.12609v4
Knowledge Distillation $\approx$ Label Smoothing: Fact or Fallacy?
Md Arafat Sultan
cs.LG, cs.CL
Originally proposed as a method for knowledge transfer from one model to another, some recent studies have suggested that knowledge distillation (KD) is in fact a form of regularization. Perhaps the strongest argument of all for this new perspective comes from its apparent similarities with label smoothing (LS). Here w...
http://arxiv.org/pdf/2301.12609v4
2023-10-25T03:10:01
2023-01-30T02:05:24
{ "title": [ -0.041534424, -0.0029945374, -0.080444336, -0.03201294, 0.0016670227, -0.02822876, -0.018096924, -0.004737854, -0.015090942, -0.0047416687, -0.024047852, -0.021728516, -0.020309448, -0.041900635, -0.016601562, -0.04071045, 0.0027599335, ...
http://arxiv.org/abs/2301.12608v2
Evaluating Neuron Interpretation Methods of NLP Models
Yimin Fan, Fahim Dalvi, Nadir Durrani, Hassan Sajjad
cs.CL
Neuron Interpretation has gained traction in the field of interpretability, and have provided fine-grained insights into what a model learns and how language knowledge is distributed amongst its different components. However, the lack of evaluation benchmark and metrics have led to siloed progress within these various ...
http://arxiv.org/pdf/2301.12608v2
2023-11-05T12:57:27
2023-01-30T02:04:35
{ "title": [ -0.020629883, 0.004211426, -0.018859863, -0.015975952, -0.012077332, -0.007347107, -0.030426025, -0.031082153, -0.009765625, 0.017715454, -0.017562866, -0.01979065, -0.018814087, -0.02319336, -0.058380127, 0.0014886856, -0.019958496, ...
http://arxiv.org/abs/2301.12596v3
Learning to Speak from Text: Zero-Shot Multilingual Text-to-Speech with Unsupervised Text Pretraining
Takaaki Saeki, Soumi Maiti, Xinjian Li, Shinji Watanabe, Shinnosuke Takamichi, Hiroshi Saruwatari
eess.AS, cs.CL
While neural text-to-speech (TTS) has achieved human-like natural synthetic speech, multilingual TTS systems are limited to resource-rich languages due to the need for paired text and studio-quality audio data. This paper proposes a method for zero-shot multilingual TTS using text-only data for the target language. The...
http://arxiv.org/pdf/2301.12596v3
2023-05-27T15:15:39
2023-01-30T00:53:50
{ "title": [ -0.006855011, -0.0073661804, -0.018249512, -0.058166504, -0.013000488, 0.012390137, -0.03765869, 0.011955261, -0.008026123, 0.044891357, -0.02267456, 0.024246216, 0.0048446655, -0.010910034, -0.034057617, -0.0011320114, -0.007091522, ...
http://arxiv.org/abs/2301.12586v2
Unifying Molecular and Textual Representations via Multi-task Language Modelling
Dimitrios Christofidellis, Giorgio Giannone, Jannis Born, Ole Winther, Teodoro Laino, Matteo Manica
cs.LG, cs.CL
The recent advances in neural language models have also been successfully applied to the field of chemistry, offering generative solutions for classical problems in molecular design and synthesis planning. These new methods have the potential to fuel a new era of data-driven automation in scientific discovery. However,...
http://arxiv.org/pdf/2301.12586v2
2023-05-18T00:37:00
2023-01-29T23:56:45
{ "title": [ -0.051605225, -0.021514893, -0.059936523, -0.068237305, -0.01977539, 0.017501831, -0.08355713, 0.032318115, 0.026519775, 0.041259766, 0.019546509, -0.00944519, 0.0155181885, -0.029083252, -0.037109375, 0.030761719, -0.005329132, 0.04...
http://arxiv.org/abs/2301.12568v1
Schema-Guided Semantic Accuracy: Faithfulness in Task-Oriented Dialogue Response Generation
Jinghong Chen, Weizhe Lin, Bill Byrne
cs.CL
Ensuring that generated utterances are faithful to dialogue actions is crucial for Task-Oriented Dialogue Response Generation. Slot Error Rate (SER) only partially measures generation quality in that it solely assesses utterances generated from non-categorical slots whose values are expected to be reproduced exactly. U...
http://arxiv.org/pdf/2301.12568v1
2023-01-29T22:32:48
2023-01-29T22:32:48
{ "title": [ -0.026046753, -0.01940918, -0.045318604, -0.05935669, -0.013725281, -0.005832672, -0.047851562, 0.018157959, -0.011528015, 0.03933716, 0.010910034, 0.0056762695, 0.002494812, -0.027801514, -0.05633545, 0.003540039, -0.0018548965, 0.0...
http://arxiv.org/abs/2301.12566v1
Improving Cross-lingual Information Retrieval on Low-Resource Languages via Optimal Transport Distillation
Zhiqi Huang, Puxuan Yu, James Allan
cs.CL, cs.IR
Benefiting from transformer-based pre-trained language models, neural ranking models have made significant progress. More recently, the advent of multilingual pre-trained language models provides great support for designing neural cross-lingual retrieval models. However, due to unbalanced pre-training data in different...
http://arxiv.org/pdf/2301.12566v1
2023-01-29T22:30:36
2023-01-29T22:30:36
{ "title": [ 0.0078048706, -0.0103302, -0.018157959, -0.070617676, 0.003107071, 0.0039405823, -0.044006348, -0.008392334, 0.033996582, 0.046722412, -0.01878357, -0.008865356, 0.016555786, -0.009262085, -0.047424316, 0.0009021759, -0.009094238, 0....
http://arxiv.org/abs/2301.12564v2
A Discerning Several Thousand Judgments: GPT-3 Rates the Article + Adjective + Numeral + Noun Construction
Kyle Mahowald
cs.CL
Knowledge of syntax includes knowledge of rare, idiosyncratic constructions. LLMs must overcome frequency biases in order to master such constructions. In this study, I prompt GPT-3 to give acceptability judgments on the English-language Article + Adjective + Numeral + Noun construction (e.g., "a lovely five days"). I ...
http://arxiv.org/pdf/2301.12564v2
2023-02-03T19:48:52
2023-01-29T22:29:55
{ "title": [ 0.033569336, 0.008605957, -0.014533997, -0.005962372, -0.0077438354, -0.023452759, -0.0045394897, -0.030639648, 0.026870728, -0.022232056, -0.0970459, 0.024169922, -0.045135498, -0.0015745163, 0.01272583, -0.057373047, -0.018295288, ...
http://arxiv.org/abs/2301.12534v4
Vicarious Offense and Noise Audit of Offensive Speech Classifiers: Unifying Human and Machine Disagreement on What is Offensive
Tharindu Cyril Weerasooriya, Sujan Dutta, Tharindu Ranasinghe, Marcos Zampieri, Christopher M. Homan, Ashiqur R. KhudaBukhsh
cs.CL, cs.CY, cs.LG
Offensive speech detection is a key component of content moderation. However, what is offensive can be highly subjective. This paper investigates how machine and human moderators disagree on what is offensive when it comes to real-world social web political discourse. We show that (1) there is extensive disagreement am...
http://arxiv.org/pdf/2301.12534v4
2023-11-09T13:46:40
2023-01-29T20:39:21
{ "title": [ 0.029708862, -0.0009698868, -0.029129028, -0.033416748, -0.008178711, -0.047454834, 0.028686523, -0.013404846, 0.020523071, 0.026428223, -0.049957275, 0.002500534, 0.004917145, -0.011734009, -0.051361084, -0.035461426, -0.014289856, ...
http://arxiv.org/abs/2301.12520v1
Producing Usable Taxonomies Cheaply and Rapidly at Pinterest Using Discovered Dynamic $μ$-Topics
Abhijit Mahabal, Jiyun Luo, Rui Huang, Michael Ellsworth, Rui Li
cs.CL
Creating a taxonomy of interests is expensive and human-effort intensive: not only do we need to identify nodes and interconnect them, in order to use the taxonomy, we must also connect the nodes to relevant entities such as users, pins, and queries. Connecting to entities is challenging because of ambiguities inherent...
http://arxiv.org/pdf/2301.12520v1
2023-01-29T19:27:08
2023-01-29T19:27:08
{ "title": [ 0.022018433, 0.000910759, -0.05581665, -0.036193848, -0.0047035217, -0.030227661, 0.00031018257, -0.01637268, -0.010627747, 0.0037784576, -0.025756836, 0.039978027, -0.0033740997, -0.026290894, 0.01576233, -0.003643036, -0.0015296936, ...
http://arxiv.org/abs/2301.13609v1
Physarum Inspired Bicycle Lane Network Design in a Congested Mega City
Md. Ahsan Habib, M. A. H. Akhand
physics.soc-ph, cs.CL
Mobility is a key factor in urban life and transport network plays a vital role in mobility. Worse transport network having less mobility is one of the key reasons to decline the living standard in any unplanned mega city. Transport mobility enhancement in an unplanned mega city is always challenging due to various con...
http://arxiv.org/pdf/2301.13609v1
2023-01-29T16:55:49
2023-01-29T16:55:49
{ "title": [ -0.01525116, -0.028884888, -0.0064086914, 0.02015686, -0.068847656, -0.020996094, -0.00491333, 0.0037631989, 0.0019836426, 0.023925781, -0.08685303, -0.007335663, 0.012908936, -0.013641357, 0.007286072, -0.044799805, 0.037628174, -0....
http://arxiv.org/abs/2301.12473v1
Large Language Models for Biomedical Causal Graph Construction
Vahan Arsenyan, Davit Shahnazaryan
cs.CL, cs.LG
Automatic causal graph construction is of high importance in medical research. They have many applications, such as clinical trial criteria design, where identification of confounding variables is a crucial step. The quality bar for clinical applications is high, and the lack of public corpora is a barrier for such stu...
http://arxiv.org/pdf/2301.12473v1
2023-01-29T15:52:33
2023-01-29T15:52:33
{ "title": [ -0.030960083, -0.029266357, -0.045562744, -0.1003418, -0.021377563, 0.025039673, -0.09362793, 0.018539429, -0.009109497, 0.009017944, 0.020004272, -0.0024871826, -0.01676941, -0.044189453, -0.009292603, 0.011222839, -0.010749817, 0.0...
http://arxiv.org/abs/2301.12463v1
Linguistic Analysis using Paninian System of Sounds and Finite State Machines
Shreekanth M Prabhu, Abhisek Midye
cs.CL, cs.FL
The study of spoken languages comprises phonology, morphology, and grammar. Analysis of a language can be based on its syntax, semantics, and pragmatics. The languages can be classified as root languages, inflectional languages, and stem languages. All these factors lead to the formation of vocabulary which has commona...
http://arxiv.org/pdf/2301.12463v1
2023-01-29T15:22:10
2023-01-29T15:22:10
{ "title": [ -0.029647827, 0.011856079, -0.048309326, -0.042633057, 0.004058838, -0.020599365, -0.03756714, -0.009521484, 0.011367798, 0.015777588, -0.01876831, -0.021453857, 0.019012451, -0.0390625, -0.027908325, 0.002527237, 0.009239197, 0.0176...
http://arxiv.org/abs/2301.12417v1
Syrupy Mouthfeel and Hints of Chocolate -- Predicting Coffee Review Scores using Text Based Sentiment
Christopher Lohse, Jeroen Lemsom, Athanasios Kalogiratos
cs.CL
This paper uses textual data contained in certified (q-graded) coffee reviews to predict corresponding scores on a scale from 0-100. By transforming this highly specialized and standardized textual data in a predictor space, we construct regression models which accurately capture the patterns in corresponding coffee be...
http://arxiv.org/pdf/2301.12417v1
2023-01-29T10:55:36
2023-01-29T10:55:36
{ "title": [ -0.021331787, -0.015220642, -0.017120361, 8.940697e-7, -0.018218994, -0.006340027, 0.013511658, -0.0035858154, -0.023345947, 0.032287598, -0.057525635, 0.0035190582, 0.013473511, -0.021469116, -0.008972168, -0.04333496, 0.0019836426, ...
http://arxiv.org/abs/2301.12376v1
Enhancing Dialogue Summarization with Topic-Aware Global- and Local- Level Centrality
Xinnian Liang, Shuangzhi Wu, Chenhao Cui, Jiaqi Bai, Chao Bian, Zhoujun Li
cs.CL
Dialogue summarization aims to condense a given dialogue into a simple and focused summary text. Typically, both the roles' viewpoints and conversational topics change in the dialogue stream. Thus how to effectively handle the shifting topics and select the most salient utterance becomes one of the major challenges of ...
http://arxiv.org/pdf/2301.12376v1
2023-01-29T06:41:55
2023-01-29T06:41:55
{ "title": [ 0.021057129, 0.007572174, -0.029724121, -0.08392334, -0.0056648254, 0.008712769, -0.06542969, -0.0025787354, 0.010566711, 0.049041748, 0.031829834, 0.03012085, 0.009132385, -0.014930725, -0.045135498, 0.031799316, -0.0140686035, 0.02...
http://arxiv.org/abs/2301.12343v1
Achieving Timestamp Prediction While Recognizing with Non-Autoregressive End-to-End ASR Model
Xian Shi, Yanni Chen, Shiliang Zhang, Zhijie Yan
cs.SD, cs.CL, eess.AS
Conventional ASR systems use frame-level phoneme posterior to conduct force-alignment~(FA) and provide timestamps, while end-to-end ASR systems especially AED based ones are short of such ability. This paper proposes to perform timestamp prediction~(TP) while recognizing by utilizing continuous integrate-and-fire~(CIF)...
http://arxiv.org/pdf/2301.12343v1
2023-01-29T03:47:59
2023-01-29T03:47:59
{ "title": [ 0.006679535, 0.026138306, -0.050964355, -0.106933594, 0.026138306, -0.0036392212, -0.054748535, 0.022781372, 0.022216797, 0.016342163, -0.01537323, -0.023223877, 0.012550354, -0.034057617, -0.054992676, -0.0046577454, -0.0023460388, ...
http://arxiv.org/abs/2301.12331v2
Time out of Mind: Generating Rate of Speech conditioned on emotion and speaker
Navjot Kaur, Paige Tuttosi
cs.CL, cs.SD, eess.AS
Voice synthesis has seen significant improvements in the past decade resulting in highly intelligible voices. Further investigations have resulted in models that can produce variable speech, including conditional emotional expression. The problem lies, however, in a focus on phrase-level modifications and prosodic voca...
http://arxiv.org/pdf/2301.12331v2
2023-01-31T05:30:55
2023-01-29T02:58:01
{ "title": [ -0.004497528, -0.000097095966, -0.04852295, -0.02784729, -0.010795593, -0.02848816, -0.00078868866, -0.0024490356, -0.023132324, 0.011428833, -0.018066406, 0.011543274, -0.0014753342, -0.041870117, -0.0087509155, -0.0138549805, -0.017059...
http://arxiv.org/abs/2301.12314v1
Progressive Prompts: Continual Learning for Language Models
Anastasia Razdaibiedina, Yuning Mao, Rui Hou, Madian Khabsa, Mike Lewis, Amjad Almahairi
cs.CL, cs.AI, cs.LG
We introduce Progressive Prompts - a simple and efficient approach for continual learning in language models. Our method allows forward transfer and resists catastrophic forgetting, without relying on data replay or a large number of task-specific parameters. Progressive Prompts learns a new soft prompt for each task a...
http://arxiv.org/pdf/2301.12314v1
2023-01-29T00:17:38
2023-01-29T00:17:38
{ "title": [ -0.011238098, -0.0027942657, -0.036315918, -0.0836792, 0.01071167, 0.0006580353, -0.057006836, 0.0062026978, 0.021865845, 0.023086548, -0.02458191, 0.0023059845, 0.00818634, -0.020462036, -0.04727173, 0.016937256, 0.0029850006, 0.025...
http://arxiv.org/abs/2301.12307v2
MQAG: Multiple-choice Question Answering and Generation for Assessing Information Consistency in Summarization
Potsawee Manakul, Adian Liusie, Mark J. F. Gales
cs.CL
State-of-the-art summarization systems can generate highly fluent summaries. These summaries, however, may contain factual inconsistencies and/or information not present in the source. Hence, an important component of assessing the quality of summaries is to determine whether there is information consistency between th...
http://arxiv.org/pdf/2301.12307v2
2023-09-07T18:20:40
2023-01-28T23:08:25
{ "title": [ -0.028182983, 0.0032291412, -0.044067383, -0.05859375, -0.0009469986, 0.008895874, -0.0395813, -0.010314941, 0.0034942627, 0.039794922, -0.0072135925, 0.02015686, -0.018829346, -0.007663727, -0.07116699, 0.0008778572, -0.013153076, 0...
http://arxiv.org/abs/2301.12303v1
Presence of informal language, such as emoticons, hashtags, and slang, impact the performance of sentiment analysis models on social media text?
Aadil Gani Ganie
cs.CL
This study aimed to investigate the influence of the presence of informal language, such as emoticons and slang, on the performance of sentiment analysis models applied to social media text. A convolutional neural network (CNN) model was developed and trained on three datasets: a sarcasm dataset, a sentiment dataset, a...
http://arxiv.org/pdf/2301.12303v1
2023-01-28T22:21:51
2023-01-28T22:21:51
{ "title": [ 0.011108398, 0.00856781, -0.010314941, 0.0146865845, -0.008964539, -0.016479492, -0.008392334, -0.033325195, 0.017028809, 0.053894043, -0.03717041, 0.002773285, 0.026519775, 0.01008606, -0.012878418, -0.012702942, 0.023498535, 0.0484...
http://arxiv.org/abs/2302.01292v1
How learners produce data from text in classifying clickbait
Nicholas J. Horton, Jie Chao, Phebe Palmer, William Finzer
cs.CY, cs.CL
Text provides a compelling example of unstructured data that can be used to motivate and explore classification problems. Challenges arise regarding the representation of features of text and student linkage between text representations as character strings and identification of features that embed connections with und...
http://arxiv.org/pdf/2302.01292v1
2023-01-28T20:23:39
2023-01-28T20:23:39
{ "title": [ 0.017578125, 0.0012245178, -0.020904541, -0.026275635, -0.006462097, -0.030853271, 0.010597229, -0.019622803, -0.025756836, 0.027450562, -0.038513184, 0.014335632, -0.010025024, -0.014930725, -0.022613525, -0.033355713, -0.017791748, ...
http://arxiv.org/abs/2301.12217v1
Semantic Parsing for Conversational Question Answering over Knowledge Graphs
Laura Perez-Beltrachini, Parag Jain, Emilio Monti, Mirella Lapata
cs.CL
In this paper, we are interested in developing semantic parsers which understand natural language questions embedded in a conversation with a user and ground them to formal queries over definitions in a general purpose knowledge graph (KG) with very large vocabularies (covering thousands of concept names and relations,...
http://arxiv.org/pdf/2301.12217v1
2023-01-28T14:45:11
2023-01-28T14:45:11
{ "title": [ -0.029434204, -0.005706787, -0.041412354, -0.079589844, -0.010597229, 0.015899658, -0.08544922, 0.01940918, 0.017288208, 0.029434204, -0.026672363, -0.0007133484, -0.004825592, 0.0026741028, -0.06982422, -0.011695862, -0.009246826, 0...
http://arxiv.org/abs/2301.12206v1
Semantic Tagging with LSTM-CRF
Farshad Noravesh
cs.CL
In the present paper, two models are presented namely LSTM-CRF and BERT-LSTM-CRF for semantic tagging of universal semantic tag dataset. The experiments show that the first model is much easier to converge while the second model that leverages BERT embedding, takes a long time to converge and needs a big dataset for se...
http://arxiv.org/pdf/2301.12206v1
2023-01-28T14:06:17
2023-01-28T14:06:17
{ "title": [ 0.0042419434, 0.007610321, -0.041870117, -0.07739258, 0.019180298, 0.0020065308, -0.04550171, -0.009391785, -0.013710022, 0.03451538, -0.01411438, 0.025558472, 0.0009841919, -0.00969696, -0.024291992, 0.022445679, -0.009178162, 0.034...
http://arxiv.org/abs/2301.12140v1
Multilingual Sentence Transformer as A Multilingual Word Aligner
Weikang Wang, Guanhua Chen, Hanqing Wang, Yue Han, Yun Chen
cs.CL
Multilingual pretrained language models (mPLMs) have shown their effectiveness in multilingual word alignment induction. However, these methods usually start from mBERT or XLM-R. In this paper, we investigate whether multilingual sentence Transformer LaBSE is a strong multilingual word aligner. This idea is non-trivial...
http://arxiv.org/pdf/2301.12140v1
2023-01-28T09:28:55
2023-01-28T09:28:55
{ "title": [ -0.02180481, 0.00035119057, -0.046417236, -0.015556335, -0.036956787, -0.010932922, -0.062286377, -0.018463135, -0.0038776398, 0.07861328, -0.005428314, -0.007282257, 0.07489014, -0.027282715, -0.029541016, 0.035003662, 0.0034122467, ...
http://arxiv.org/abs/2301.12139v3
Bipol: Multi-axes Evaluation of Bias with Explainability in Benchmark Datasets
Tosin Adewumi, Isabella Södergren, Lama Alkhaled, Sana Sabah Sabry, Foteini Liwicki, Marcus Liwicki
cs.CL
We investigate five English NLP benchmark datasets (on the superGLUE leaderboard) and two Swedish datasets for bias, along multiple axes. The datasets are the following: Boolean Question (Boolq), CommitmentBank (CB), Winograd Schema Challenge (WSC), Wino-gender diagnostic (AXg), Recognising Textual Entailment (RTE), Sw...
http://arxiv.org/pdf/2301.12139v3
2023-09-16T15:56:11
2023-01-28T09:28:19
{ "title": [ -0.02470398, -0.0020103455, -0.035491943, -0.06149292, 0.014770508, 0.0049095154, -0.041168213, -0.019226074, -0.00051403046, -0.007621765, -0.028823853, 0.011108398, -0.027893066, -0.034942627, -0.017669678, -0.021942139, -0.037597656, ...
http://arxiv.org/abs/2301.12134v1
Underwater Robotics Semantic Parser Assistant
Parth Parekh, Cedric McGuire, Jake Imyak
cs.CL, cs.AI, I.2.7; C.3
Semantic parsing is a means of taking natural language and putting it in a form that a computer can understand. There has been a multitude of approaches that take natural language utterances and form them into lambda calculus expressions -- mathematical functions to describe logic. Here, we experiment with a sequence t...
http://arxiv.org/pdf/2301.12134v1
2023-01-28T09:04:43
2023-01-28T09:04:43
{ "title": [ -0.02015686, 0.0028839111, -0.059783936, -0.019470215, -0.0143585205, -0.022628784, -0.03375244, -0.009239197, -0.020874023, 0.043640137, -0.055267334, -0.0038452148, 0.04067993, -0.022003174, -0.010421753, -0.023956299, -0.0075149536, ...
http://arxiv.org/abs/2301.12132v2
AutoPEFT: Automatic Configuration Search for Parameter-Efficient Fine-Tuning
Han Zhou, Xingchen Wan, Ivan Vulić, Anna Korhonen
cs.CL, cs.AI, cs.LG
Large pretrained language models are widely used in downstream NLP tasks via task-specific fine-tuning, but such procedures can be costly. Recently, Parameter-Efficient Fine-Tuning (PEFT) methods have achieved strong task performance while updating a much smaller number of parameters compared to full model fine-tuning ...
http://arxiv.org/pdf/2301.12132v2
2023-06-06T17:07:23
2023-01-28T08:51:23
{ "title": [ -0.03439331, -0.035583496, -0.027038574, -0.093322754, 0.003030777, 0.016082764, -0.05429077, 0.012619019, -0.0118637085, 0.020889282, 0.01020813, 0.006439209, -0.009254456, -0.033996582, -0.022628784, 0.0058135986, -0.0065727234, -0...
http://arxiv.org/abs/2301.12112v2
On Pre-trained Language Models for Antibody
Danqing Wang, Fei Ye, Hao Zhou
cs.CL, q-bio.BM
Antibodies are vital proteins offering robust protection for the human body from pathogens. The development of general protein and antibody-specific pre-trained language models both facilitate antibody prediction tasks. However, there have been limited studies that comprehensively explore the representation capability ...
http://arxiv.org/pdf/2301.12112v2
2023-03-01T20:18:25
2023-01-28T07:05:15
{ "title": [ -0.013946533, -0.0013465881, -0.023529053, -0.051116943, -0.018127441, 0.0061302185, -0.056427002, 0.0019798279, 0.00025105476, 0.03817749, 0.008155823, -0.016693115, 0.017669678, -0.0413208, -0.025634766, 0.017425537, 0.010673523, 0...
http://arxiv.org/abs/2302.10983v1
Do Orcas Have Semantic Language? Machine Learning to Predict Orca Behaviors Using Partially Labeled Vocalization Data
Sophia Sandholm
cs.SD, cs.CL, cs.LG, eess.AS
Orcinus orca (killer whales) exhibit complex calls. They last about a second. In a call, an orca typically uses multiple frequencies simultaneously, varies the frequencies, and varies their volumes. Behavior data is hard to obtain because orcas live under water and travel quickly. Sound data is relatively easy to captu...
http://arxiv.org/pdf/2302.10983v1
2023-01-28T06:04:22
2023-01-28T06:04:22
{ "title": [ -0.036987305, 0.012435913, -0.04800415, -0.0013656616, -0.019561768, -0.029769897, -0.018432617, -0.014854431, -0.035186768, 0.022949219, -0.031677246, -0.002714157, -0.01838684, -0.04107666, 0.013824463, -0.04827881, 0.0115737915, 0...
http://arxiv.org/abs/2301.12074v1
Comparing Intrinsic Gender Bias Evaluation Measures without using Human Annotated Examples
Masahiro Kaneko, Danushka Bollegala, Naoaki Okazaki
cs.CL
Numerous types of social biases have been identified in pre-trained language models (PLMs), and various intrinsic bias evaluation measures have been proposed for quantifying those social biases. Prior works have relied on human annotated examples to compare existing intrinsic bias evaluation measures. However, this app...
http://arxiv.org/pdf/2301.12074v1
2023-01-28T03:11:50
2023-01-28T03:11:50
{ "title": [ -0.021133423, 0.00920105, -0.056549072, -0.0076446533, 0.013885498, -0.035369873, 0.014411926, -0.024795532, -0.0043296814, 0.040161133, -0.040618896, 0.000046014786, -0.017715454, 0.0037174225, -0.057159424, -0.018035889, -0.028533936, ...
http://arxiv.org/abs/2301.12050v2
Do Embodied Agents Dream of Pixelated Sheep: Embodied Decision Making using Language Guided World Modelling
Kolby Nottingham, Prithviraj Ammanabrolu, Alane Suhr, Yejin Choi, Hannaneh Hajishirzi, Sameer Singh, Roy Fox
cs.LG, cs.CL
Reinforcement learning (RL) agents typically learn tabula rasa, without prior knowledge of the world. However, if initialized with knowledge of high-level subgoals and transitions between subgoals, RL agents could utilize this Abstract World Model (AWM) for planning and exploration. We propose using few-shot large lang...
http://arxiv.org/pdf/2301.12050v2
2023-04-27T15:14:01
2023-01-28T02:04:07
{ "title": [ -0.0309906, -0.02696228, 0.0022697449, 0.012001038, -0.05203247, -0.002105713, 0.007083893, -0.022583008, -0.014892578, 0.026977539, -0.040008545, 0.013298035, 0.00932312, -0.01864624, -0.01751709, -0.017303467, -0.006214142, 0.03089...
http://arxiv.org/abs/2301.12017v2
Understanding INT4 Quantization for Transformer Models: Latency Speedup, Composability, and Failure Cases
Xiaoxia Wu, Cheng Li, Reza Yazdani Aminabadi, Zhewei Yao, Yuxiong He
cs.CL, cs.LG
Improving the deployment efficiency of transformer-based language models has been challenging given their high computation and memory cost. While INT8 quantization has recently been shown to be effective in reducing both the memory cost and latency while preserving model accuracy, it remains unclear whether we can leve...
http://arxiv.org/pdf/2301.12017v2
2023-05-30T21:32:11
2023-01-27T22:44:18
{ "title": [ -0.017852783, 0.009490967, -0.0105896, -0.014343262, -0.013435364, -0.0053749084, -0.019683838, -0.031188965, -0.015144348, 0.07543945, 0.0045433044, -0.009117126, 0.04159546, -0.02645874, -0.040802002, 0.012321472, 0.009674072, 0.02...
http://arxiv.org/abs/2301.12006v1
Improved knowledge distillation by utilizing backward pass knowledge in neural networks
Aref Jafari, Mehdi Rezagholizadeh, Ali Ghodsi
cs.LG, cs.CL, cs.CV
Knowledge distillation (KD) is one of the prominent techniques for model compression. In this method, the knowledge of a large network (teacher) is distilled into a model (student) with usually significantly fewer parameters. KD tries to better-match the output of the student model to that of the teacher model based on...
http://arxiv.org/pdf/2301.12006v1
2023-01-27T22:07:38
2023-01-27T22:07:38
{ "title": [ -0.0051574707, 0.01625061, -0.051757812, -0.072387695, 0.027389526, -0.012916565, -0.046203613, -0.019424438, 0.033935547, 0.023345947, -0.037719727, -0.01499176, -0.026794434, -0.01838684, -0.061401367, -0.015556335, 0.010276794, 0....
http://arxiv.org/abs/2301.12004v1
Understanding the Effectiveness of Very Large Language Models on Dialog Evaluation
Jessica Huynh, Cathy Jiao, Prakhar Gupta, Shikib Mehri, Payal Bajaj, Vishrav Chaudhary, Maxine Eskenazi
cs.CL
Language models have steadily increased in size over the past few years. They achieve a high level of performance on various natural language processing (NLP) tasks such as question answering and summarization. Large language models (LLMs) have been used for generation and can now output human-like text. Due to this, t...
http://arxiv.org/pdf/2301.12004v1
2023-01-27T22:02:27
2023-01-27T22:02:27
{ "title": [ -0.0054130554, -0.0034313202, -0.04046631, -0.066223145, 0.0011672974, -0.013839722, -0.056396484, -0.00333786, 0.00995636, 0.032165527, -0.0012922287, -0.0132751465, -0.0067596436, -0.022476196, -0.06286621, 0.00055360794, -0.024490356,...
http://arxiv.org/abs/2301.11997v1
Prompt-Based Editing for Text Style Transfer
Guoqing Luo, Yu Tong Han, Lili Mou, Mauajama Firdaus
cs.CL, cs.AI, cs.LG
Prompting approaches have been recently explored in text style transfer, where a textual prompt is used to query a pretrained language model to generate style-transferred texts word by word in an autoregressive manner. However, such a generation process is less controllable and early prediction errors may affect future...
http://arxiv.org/pdf/2301.11997v1
2023-01-27T21:31:14
2023-01-27T21:31:14
{ "title": [ -0.02909851, -0.007347107, -0.03652954, -0.03289795, -0.011695862, -0.0001320839, -0.028625488, -0.010658264, 0.000869751, 0.05053711, -0.006439209, 0.011795044, 0.026260376, -0.018554688, -0.028793335, 0.014297485, 0.003835678, 0.00...
http://arxiv.org/abs/2301.11916v3
Large Language Models Are Latent Variable Models: Explaining and Finding Good Demonstrations for In-Context Learning
Xinyi Wang, Wanrong Zhu, Michael Saxon, Mark Steyvers, William Yang Wang
cs.CL, cs.AI, cs.LG
In recent years, pre-trained large language models (LLMs) have demonstrated remarkable efficiency in achieving an inference-time few-shot learning capability known as in-context learning. However, existing literature has highlighted the sensitivity of this capability to the selection of few-shot demonstrations. Current...
http://arxiv.org/pdf/2301.11916v3
2023-10-17T11:24:33
2023-01-27T18:59:01
{ "title": [ -0.02268982, -0.0082473755, -0.038146973, -0.048950195, -0.018447876, -0.028396606, -0.035705566, -0.00007635355, -0.02432251, 0.009307861, -0.045410156, 0.020111084, -0.017547607, -0.024368286, 0.008583069, -0.022460938, -0.009086609, ...
http://arxiv.org/abs/2301.11879v2
Case-Based Reasoning with Language Models for Classification of Logical Fallacies
Zhivar Sourati, Filip Ilievski, Hông-Ân Sandlin, Alain Mermoud
cs.AI, cs.CL
The ease and speed of spreading misinformation and propaganda on the Web motivate the need to develop trustworthy technology for detecting fallacies in natural language arguments. However, state-of-the-art language modeling methods exhibit a lack of robustness on tasks like logical fallacy classification that require c...
http://arxiv.org/pdf/2301.11879v2
2023-05-17T20:13:59
2023-01-27T17:49:16
{ "title": [ 0.023239136, -0.014297485, -0.025222778, -0.0690918, 0.00042009354, -0.0085372925, -0.03753662, -0.0075416565, -0.004135132, 0.03250122, -0.025848389, 0.004020691, -0.005542755, -0.015319824, -0.04043579, -0.0128479, -0.03201294, 0.0...
http://arxiv.org/abs/2301.11850v3
Predicting Sentence-Level Factuality of News and Bias of Media Outlets
Francielle Vargas, Kokil Jaidka, Thiago A. S. Pardo, Fabrício Benevenuto
cs.CL
Automated news credibility and fact-checking at scale require accurately predicting news factuality and media bias. This paper introduces a large sentence-level dataset, titled "FactNews", composed of 6,191 sentences expertly annotated according to factuality and media bias definitions proposed by AllSides. We use Fact...
http://arxiv.org/pdf/2301.11850v3
2023-06-28T21:11:39
2023-01-27T16:56:24
{ "title": [ 0.0038852692, 0.0013074875, -0.04623413, -0.04449463, -0.0029506683, -0.03265381, -0.0082092285, -0.023422241, -0.016448975, 0.033996582, -0.035339355, 0.01876831, 0.0206604, -0.004184723, -0.027908325, -0.0019989014, -0.016906738, 0...
http://arxiv.org/abs/2301.11847v1
A Comparative Study of Pretrained Language Models for Long Clinical Text
Yikuan Li, Ramsey M. Wehbe, Faraz S. Ahmad, Hanyin Wang, Yuan Luo
cs.CL, cs.AI
Objective: Clinical knowledge enriched transformer models (e.g., ClinicalBERT) have state-of-the-art results on clinical NLP (natural language processing) tasks. One of the core limitations of these transformer models is the substantial memory consumption due to their full self-attention mechanism, which leads to the p...
http://arxiv.org/pdf/2301.11847v1
2023-01-27T16:50:29
2023-01-27T16:50:29
{ "title": [ -0.017150879, -0.0053901672, -0.03173828, -0.07476807, -0.015159607, 0.0042266846, -0.08679199, -0.0007953644, -0.014137268, 0.0423584, 0.004295349, 0.010620117, 0.0038604736, -0.023406982, -0.03387451, 0.0060424805, -0.009605408, 0....
http://arxiv.org/abs/2301.11845v2
Learning the Effects of Physical Actions in a Multi-modal Environment
Gautier Dagan, Frank Keller, Alex Lascarides
cs.CL
Large Language Models (LLMs) handle physical commonsense information inadequately. As a result of being trained in a disembodied setting, LLMs often fail to predict an action's outcome in a given environment. However, predicting the effects of an action before it is executed is crucial in planning, where coherent seque...
http://arxiv.org/pdf/2301.11845v2
2023-02-03T12:29:30
2023-01-27T16:49:52
{ "title": [ 0.030670166, 0.016403198, -0.026535034, -0.035125732, -0.001086235, -0.045410156, -0.025390625, -0.033966064, -0.018493652, 0.026611328, -0.03152466, 0.012374878, 0.0071372986, -0.025741577, -0.0021419525, -0.017333984, 0.013267517, ...
http://arxiv.org/abs/2301.11843v1
Reading and Reasoning over Chart Images for Evidence-based Automated Fact-Checking
Mubashara Akhtar, Oana Cocarascu, Elena Simperl
cs.CL, cs.CV
Evidence data for automated fact-checking (AFC) can be in multiple modalities such as text, tables, images, audio, or video. While there is increasing interest in using images for AFC, previous works mostly focus on detecting manipulated or fake images. We propose a novel task, chart-based fact-checking, and introduce ...
http://arxiv.org/pdf/2301.11843v1
2023-01-27T16:47:45
2023-01-27T16:47:45
{ "title": [ 0.018676758, -0.006916046, -0.020767212, -0.062683105, -0.0030403137, -0.010032654, -0.017486572, -0.017166138, 0.0014047623, 0.031311035, -0.024368286, -0.00010949373, -0.0041236877, -0.019699097, -0.039642334, -0.005844116, 0.000638484...
http://arxiv.org/abs/2301.11796v1
Call for Papers -- The BabyLM Challenge: Sample-efficient pretraining on a developmentally plausible corpus
Alex Warstadt, Leshem Choshen, Aaron Mueller, Adina Williams, Ethan Wilcox, Chengxu Zhuang
cs.CL
We present the call for papers for the BabyLM Challenge: Sample-efficient pretraining on a developmentally plausible corpus. This shared task is intended for participants with an interest in small scale language modeling, human language acquisition, low-resource NLP, and cognitive modeling. In partnership with CoNLL an...
http://arxiv.org/pdf/2301.11796v1
2023-01-27T15:52:50
2023-01-27T15:52:50
{ "title": [ 0.010795593, -0.013442993, -0.014289856, -0.08770752, -0.010597229, 0.008529663, -0.064086914, 0.0031719208, -0.021377563, 0.035369873, -0.01727295, 0.030731201, 0.006801605, -0.021469116, 0.0021915436, -0.0020065308, -0.014633179, 0...
http://arxiv.org/abs/2301.11792v1
Graph Attention with Hierarchies for Multi-hop Question Answering
Yunjie He, Philip John Gorinski, Ieva Staliunaite, Pontus Stenetorp
cs.CL, cs.AI
Multi-hop QA (Question Answering) is the task of finding the answer to a question across multiple documents. In recent years, a number of Deep Learning-based approaches have been proposed to tackle this complex task, as well as a few standard benchmarks to assess models Multi-hop QA capabilities. In this paper, we focu...
http://arxiv.org/pdf/2301.11792v1
2023-01-27T15:49:50
2023-01-27T15:49:50
{ "title": [ -0.0024375916, -0.009132385, -0.02571106, -0.056610107, -0.009109497, 0.00047039986, -0.030944824, 0.020462036, 0.02758789, 0.044647217, -0.026016235, 0.005874634, 0.021514893, -0.000330925, -0.04928589, 0.010429382, -0.005405426, 0....
http://arxiv.org/abs/2301.11757v3
Moûsai: Text-to-Music Generation with Long-Context Latent Diffusion
Flavio Schneider, Ojasv Kamal, Zhijing Jin, Bernhard Schölkopf
cs.CL, cs.LG, cs.SD, eess.AS
Recent years have seen the rapid development of large generative models for text; however, much less research has explored the connection between text and another "language" of communication -- music. Music, much like text, can convey emotions, stories, and ideas, and has its own unique structure and syntax. In our wor...
http://arxiv.org/pdf/2301.11757v3
2023-10-23T20:47:30
2023-01-27T14:52:53
{ "title": [ -0.02658081, -0.018447876, -0.060821533, -0.055603027, 0.005092621, -0.007949829, -0.015167236, 0.02909851, 0.004486084, 0.043395996, -0.04083252, 0.039154053, 0.017456055, -0.0075416565, -0.024246216, 0.008605957, -0.024597168, 0.04...
http://arxiv.org/abs/2301.11749v1
A Multi-task Multi-stage Transitional Training Framework for Neural Chat Translation
Chulun Zhou, Yunlong Liang, Fandong Meng, Jie Zhou, Jinan Xu, Hongji Wang, Min Zhang, Jinsong Su
cs.CL
Neural chat translation (NCT) aims to translate a cross-lingual chat between speakers of different languages. Existing context-aware NMT models cannot achieve satisfactory performances due to the following inherent problems: 1) limited resources of annotated bilingual dialogues; 2) the neglect of modelling conversation...
http://arxiv.org/pdf/2301.11749v1
2023-01-27T14:41:16
2023-01-27T14:41:16
{ "title": [ -0.021331787, -0.0082473755, -0.008895874, -0.05697632, -0.01890564, 0.021362305, -0.059509277, 0.008430481, -0.024520874, 0.064086914, 0.0104599, -0.0050621033, 0.02015686, -0.02142334, -0.049987793, 0.014167786, -0.002084732, 0.011...
http://arxiv.org/abs/2301.11719v4
The Exploration of Knowledge-Preserving Prompts for Document Summarisation
Chen Chen, Wei Emma Zhang, Alireza Seyed Shakeri, Makhmoor Fiza
cs.CL, cs.AI, cs.LG
Despite the great development of document summarisation techniques nowadays, factual inconsistencies between the generated summaries and the original texts still occur from time to time. This study explores the possibility of adopting prompts to incorporate factual knowledge into generated summaries. We specifically st...
http://arxiv.org/pdf/2301.11719v4
2023-05-17T01:58:11
2023-01-27T14:05:12
{ "title": [ -0.009223938, -0.012214661, -0.03564453, -0.05178833, -0.018722534, -0.008766174, -0.044769287, 0.003276825, 0.014434814, 0.04660034, -0.02319336, -0.0042686462, 0.015686035, -0.00040864944, -0.072143555, -0.00207901, 0.011741638, 0....
http://arxiv.org/abs/2301.11716v3
Pre-training for Speech Translation: CTC Meets Optimal Transport
Phuong-Hang Le, Hongyu Gong, Changhan Wang, Juan Pino, Benjamin Lecouteux, Didier Schwab
cs.CL, cs.LG, cs.SD, eess.AS
The gap between speech and text modalities is a major challenge in speech-to-text translation (ST). Different methods have been proposed to reduce this gap, but most of them require architectural changes in ST training. In this work, we propose to mitigate this issue at the pre-training stage, requiring no change in th...
http://arxiv.org/pdf/2301.11716v3
2023-06-05T11:44:02
2023-01-27T14:03:09
{ "title": [ -0.017196655, -0.02192688, 0.00024187565, -0.051727295, -0.04135132, 0.012069702, -0.054504395, 0.0005030632, -0.03692627, 0.050872803, -0.009689331, 0.015808105, 0.012001038, -0.028137207, -0.016952515, -0.005756378, 0.0038871765, 0...
http://arxiv.org/abs/2301.11709v1
Semantic Network Model for Sign Language Comprehension
Xinchen Kang, Dengfeng Yao, Minghu Jiang, Yunlong Huang, Fanshu Li
cs.CL
In this study, the authors propose a computational cognitive model for sign language (SL) perception and comprehension with detailed algorithmic descriptions based on cognitive functionalities in human language processing. The semantic network model (SNM) that represents semantic relations between concepts, it is used ...
http://arxiv.org/pdf/2301.11709v1
2023-01-27T13:43:02
2023-01-27T13:43:02
{ "title": [ -0.061065674, -0.008117676, -0.036743164, -0.040771484, -0.0435791, 0.018798828, -0.09222412, -0.0053520203, 0.023635864, 0.010253906, -0.053619385, -0.01663208, -0.019622803, -0.015327454, -0.052520752, -0.026824951, -0.013824463, 0...
http://arxiv.org/abs/2301.11696v1
SLCNN: Sentence-Level Convolutional Neural Network for Text Classification
Ali Jarrahi, Ramin Mousa, Leila Safari
cs.CL, cs.LG
Text classification is a fundamental task in natural language processing (NLP). Several recent studies show the success of deep learning on text processing. Convolutional neural network (CNN), as a popular deep learning model, has shown remarkable success in the task of text classification. In this paper, new baseline ...
http://arxiv.org/pdf/2301.11696v1
2023-01-27T13:16:02
2023-01-27T13:16:02
{ "title": [ -0.015853882, -0.010467529, -0.040283203, -0.05355835, 0.014602661, 0.0027866364, -0.02760315, -0.01676941, 0.0112838745, 0.024932861, -0.0104904175, 0.019042969, -0.0079574585, -0.009689331, -0.050109863, 0.00022757053, -0.019104004, ...
http://arxiv.org/abs/2301.11688v1
Can We Use Probing to Better Understand Fine-tuning and Knowledge Distillation of the BERT NLU?
Jakub Hościłowicz, Marcin Sowański, Piotr Czubowski, Artur Janicki
cs.CL, cs.LG
In this article, we use probing to investigate phenomena that occur during fine-tuning and knowledge distillation of a BERT-based natural language understanding (NLU) model. Our ultimate purpose was to use probing to better understand practical production problems and consequently to build better NLU models. We designe...
http://arxiv.org/pdf/2301.11688v1
2023-01-27T12:56:29
2023-01-27T12:56:29
{ "title": [ 0.0013093948, -0.039123535, -0.0026016235, -0.0758667, -0.036346436, 0.009864807, -0.030441284, 0.04345703, -0.02960205, 0.02709961, -0.035949707, -0.0072135925, 0.014930725, -0.031463623, 0.015792847, -0.043273926, 0.012939453, -0.0...
http://arxiv.org/abs/2301.11682v1
Towards Personalized Review Summarization by Modeling Historical Reviews from Customer and Product Separately
Xin Cheng, Shen Gao, Yuchi Zhang, Yongliang Wang, Xiuying Chen, Mingzhe Li, Dongyan Zhao, Rui Yan
cs.CL, cs.AI, cs.IR
Review summarization is a non-trivial task that aims to summarize the main idea of the product review in the E-commerce website. Different from the document summary which only needs to focus on the main facts described in the document, review summarization should not only summarize the main aspects mentioned in the rev...
http://arxiv.org/pdf/2301.11682v1
2023-01-27T12:32:55
2023-01-27T12:32:55
{ "title": [ -0.010513306, -0.005012512, -0.020080566, -0.058532715, -0.02154541, 0.0031814575, -0.042755127, -0.0046157837, -0.012039185, 0.055511475, -0.028030396, 0.026916504, 0.007095337, 0.008598328, -0.057922363, 0.0026607513, 0.0030879974, ...
http://arxiv.org/abs/2301.11660v4
Probing Out-of-Distribution Robustness of Language Models with Parameter-Efficient Transfer Learning
Hyunsoo Cho, Choonghyun Park, Junyeop Kim, Hyuhng Joon Kim, Kang Min Yoo, Sang-goo Lee
cs.CL
As the size of the pre-trained language model (PLM) continues to increase, numerous parameter-efficient transfer learning methods have been proposed recently to compensate for the tremendous cost of fine-tuning. Despite the impressive results achieved by large pre-trained language models (PLMs) and various parameter-ef...
http://arxiv.org/pdf/2301.11660v4
2023-06-14T03:12:50
2023-01-27T11:27:40
{ "title": [ -0.005580902, 0.0073928833, -0.027420044, -0.045196533, 0.014434814, -0.008636475, -0.035064697, -0.017684937, 0.004634857, 0.020401001, -0.014228821, 0.020202637, 0.0025081635, -0.017028809, -0.03366089, 0.003982544, -0.021377563, 0...
http://arxiv.org/abs/2301.11621v1
Event Causality Extraction with Event Argument Correlations
Shiyao Cui, Jiawei Sheng, Xin Cong, QuanGang Li, Tingwen Liu, Jinqiao Shi
cs.CL
Event Causality Identification (ECI), which aims to detect whether a causality relation exists between two given textual events, is an important task for event causality understanding. However, the ECI task ignores crucial event structure and cause-effect causality component information, making it struggle for downstre...
http://arxiv.org/pdf/2301.11621v1
2023-01-27T09:48:31
2023-01-27T09:48:31
{ "title": [ 0.009994507, 0.015914917, -0.047180176, -0.10455322, 0.034698486, 0.008781433, -0.055511475, -0.021072388, -0.005302429, 0.0143966675, 0.0368042, 0.0075798035, -0.009399414, -0.037994385, -0.01902771, 0.014099121, -0.018829346, 0.042...
http://arxiv.org/abs/2301.11608v1
A Multi-View Joint Learning Framework for Embedding Clinical Codes and Text Using Graph Neural Networks
Lecheng Kong, Christopher King, Bradley Fritz, Yixin Chen
cs.CL, cs.AI
Learning to represent free text is a core task in many clinical machine learning (ML) applications, as clinical text contains observations and plans not otherwise available for inference. State-of-the-art methods use large language models developed with immense computational resources and training data; however, applyi...
http://arxiv.org/pdf/2301.11608v1
2023-01-27T09:19:03
2023-01-27T09:19:03
{ "title": [ -0.02633667, -0.0053634644, -0.025527954, -0.047576904, -0.015716553, 0.011276245, -0.054779053, 0.0015563965, -0.019424438, 0.037963867, 0.0070495605, 0.008453369, -0.015670776, -0.038116455, -0.016311646, -0.00349617, -0.015960693, ...
http://arxiv.org/abs/2301.11596v5
ThoughtSource: A central hub for large language model reasoning data
Simon Ott, Konstantin Hebenstreit, Valentin Liévin, Christoffer Egeberg Hother, Milad Moradi, Maximilian Mayrhauser, Robert Praas, Ole Winther, Matthias Samwald
cs.CL, cs.AI
Large language models (LLMs) such as GPT-4 have recently demonstrated impressive results across a wide range of tasks. LLMs are still limited, however, in that they frequently fail at complex reasoning, their reasoning processes are opaque, they are prone to 'hallucinate' facts, and there are concerns about their under...
http://arxiv.org/pdf/2301.11596v5
2023-07-27T09:37:35
2023-01-27T08:45:53
{ "title": [ -0.020324707, -0.040649414, -0.045654297, -0.09301758, -0.014060974, 0.014022827, -0.076660156, 0.027709961, 0.014701843, 0.022979736, -0.014419556, -0.02053833, 0.0154418945, -0.01663208, -0.044677734, 0.013114929, -0.0038986206, 0....
http://arxiv.org/abs/2301.11579v1
Down the Rabbit Hole: Detecting Online Extremism, Radicalisation, and Politicised Hate Speech
Jarod Govers, Philip Feldman, Aaron Dant, Panos Patros
cs.SI, cs.AI, cs.CL, cs.CY, cs.LG, I.2.1; I.2.4; I.2.6; I.2.7; I.2.10; K.4.1; I.5.1; I.5.4; I.5.2; I.5.3; J.4
Social media is a modern person's digital voice to project and engage with new ideas and mobilise communities $\unicode{x2013}$ a power shared with extremists. Given the societal risks of unvetted content-moderating algorithms for Extremism, Radicalisation, and Hate speech (ERH) detection, responsible software engineer...
http://arxiv.org/pdf/2301.11579v1
2023-01-27T07:59:31
2023-01-27T07:59:31
{ "title": [ 0.029052734, 0.016143799, -0.03656006, -0.02810669, 0.036315918, -0.044708252, 0.032440186, -0.010940552, -0.023635864, 0.014343262, -0.010543823, 0.016342163, 0.02255249, -0.042877197, -0.008148193, -0.014785767, -0.01436615, 0.0605...
http://arxiv.org/abs/2301.11564v1
Learning 6-DoF Fine-grained Grasp Detection Based on Part Affordance Grounding
Yaoxian Song, Penglei Sun, Yi Ren, Yu Zheng, Yue Zhang
cs.RO, cs.CL, cs.CV, cs.HC
Robotic grasping is a fundamental ability for a robot to interact with the environment. Current methods focus on how to obtain a stable and reliable grasping pose in object wise, while little work has been studied on part (shape)-wise grasping which is related to fine-grained grasping and robotic affordance. Parts can ...
http://arxiv.org/pdf/2301.11564v1
2023-01-27T07:00:54
2023-01-27T07:00:54
{ "title": [ -0.0011081696, 0.03488159, -0.0574646, -0.043945312, 0.036712646, -0.028442383, 0.010444641, 0.004196167, 0.0013303757, 0.03050232, -0.058685303, 0.019210815, -0.018844604, 0.0053977966, -0.0496521, -0.026794434, 0.006729126, 0.03924...
http://arxiv.org/abs/2301.11508v2
Theme-driven Keyphrase Extraction to Analyze Social Media Discourse
William Romano, Omar Sharif, Madhusudan Basak, Joseph Gatto, Sarah Preum
cs.CL
Social media platforms are vital resources for sharing self-reported health experiences, offering rich data on various health topics. Despite advancements in Natural Language Processing (NLP) enabling large-scale social media data analysis, a gap remains in applying keyphrase extraction to health-related content. Keyph...
http://arxiv.org/pdf/2301.11508v2
2023-05-28T20:06:35
2023-01-27T03:00:46
{ "title": [ -0.01146698, 0.0009598732, -0.02658081, -0.06109619, 0.008483887, 0.012413025, -0.027908325, 0.008262634, 0.0017061234, 0.049713135, 0.0054359436, 0.025436401, -0.008956909, -0.00062417984, -0.060791016, 0.006038666, -0.03378296, 0.0...
http://arxiv.org/abs/2301.11507v1
Semi-Parametric Video-Grounded Text Generation
Sungdong Kim, Jin-Hwa Kim, Jiyoung Lee, Minjoon Seo
cs.CV, cs.CL, cs.LG
Efficient video-language modeling should consider the computational cost because of a large, sometimes intractable, number of video frames. Parametric approaches such as the attention mechanism may not be ideal since its computational cost quadratically increases as the video length increases. Rather, previous studies ...
http://arxiv.org/pdf/2301.11507v1
2023-01-27T03:00:43
2023-01-27T03:00:43
{ "title": [ -0.021560669, 0.007865906, -0.032928467, -0.054473877, -0.017471313, -0.000852108, -0.05368042, -0.010856628, -0.0064201355, 0.049713135, -0.034179688, 0.0021152496, 0.022537231, -0.0069732666, -0.03942871, -0.016113281, -0.00907135, ...
http://arxiv.org/abs/2301.11503v1
Candidate Soups: Fusing Candidate Results Improves Translation Quality for Non-Autoregressive Translation
Huanran Zheng, Wei Zhu, Pengfei Wang, Xiaoling Wang
cs.CL
Non-autoregressive translation (NAT) model achieves a much faster inference speed than the autoregressive translation (AT) model because it can simultaneously predict all tokens during inference. However, its translation quality suffers from degradation compared to AT. And existing NAT methods only focus on improving t...
http://arxiv.org/pdf/2301.11503v1
2023-01-27T02:39:42
2023-01-27T02:39:42
{ "title": [ -0.029144287, -0.0004544258, -0.00982666, -0.057922363, -0.015991211, 0.0038204193, -0.06738281, -0.01625061, -0.010864258, 0.041748047, 0.00856781, 0.027954102, 0.01676941, -0.02909851, -0.0345459, 0.019195557, -0.0079193115, 0.0138...
http://arxiv.org/abs/2302.08901v1
Exploring External Knowledge for Accurate modeling of Visual and Language Problems
Xuewen Yang
cs.CV, cs.CL
The interest in Artificial Intelligence (AI) and its applications has seen unprecedented growth in the last few years. The success can be partly attributed to the advancements of deep neural networks made in the sub-fields of AI such as Computer Vision (CV) and Natural Language Processing (NLP). The promising research ...
http://arxiv.org/pdf/2302.08901v1
2023-01-27T02:01:50
2023-01-27T02:01:50
{ "title": [ -0.0090408325, -0.023742676, -0.009101868, -0.036254883, -0.010681152, -0.0099487305, -0.02758789, -0.01574707, 0.011993408, 0.01335144, -0.022705078, -0.010444641, -0.025268555, -0.019592285, -0.03933716, -0.019638062, 0.012290955, ...
http://arxiv.org/abs/2301.11489v2
Talk the Walk: Synthetic Data Generation for Conversational Music Recommendation
Megan Leszczynski, Ravi Ganti, Shu Zhang, Krisztian Balog, Filip Radlinski, Fernando Pereira, Arun Tejasvi Chaganty
cs.IR, cs.CL
Recommendation systems are ubiquitous yet often difficult for users to control and adjust when recommendation quality is poor. This has motivated the development of conversational recommendation systems (CRSs), with control over recommendations provided through natural language feedback. However, building conversationa...
http://arxiv.org/pdf/2301.11489v2
2023-07-08T21:41:23
2023-01-27T01:54:16
{ "title": [ -0.0026798248, -0.01626587, -0.05722046, -0.06451416, -0.004047394, -0.02078247, -0.030807495, 0.03591919, -0.022979736, 0.02168274, -0.037902832, 0.028045654, -0.027938843, -0.014854431, -0.029647827, -0.02053833, -0.025253296, 0.02...
http://arxiv.org/abs/2301.11462v2
How poor is the stimulus? Evaluating hierarchical generalization in neural networks trained on child-directed speech
Aditya Yedetore, Tal Linzen, Robert Frank, R. Thomas McCoy
cs.CL, J.4; I.2.7
When acquiring syntax, children consistently choose hierarchical rules over competing non-hierarchical possibilities. Is this preference due to a learning bias for hierarchical structure, or due to more general biases that interact with hierarchical cues in children's linguistic input? We explore these possibilities by...
http://arxiv.org/pdf/2301.11462v2
2023-06-06T13:40:22
2023-01-26T23:24:17
{ "title": [ -0.0012207031, 0.00969696, -0.028778076, -0.06890869, 0.010223389, -0.016815186, -0.022415161, 0.000521183, -0.027328491, 0.04815674, -0.009811401, 0.01889038, -0.02015686, -0.010894775, -0.054473877, -0.023849487, -0.02243042, 0.044...
http://arxiv.org/abs/2301.11459v1
Neural-Symbolic Inference for Robust Autoregressive Graph Parsing via Compositional Uncertainty Quantification
Zi Lin, Jeremiah Liu, Jingbo Shang
cs.CL
Pre-trained seq2seq models excel at graph semantic parsing with rich annotated data, but generalize worse to out-of-distribution (OOD) and long-tail examples. In comparison, symbolic parsers under-perform on population-level metrics, but exhibit unique strength in OOD and tail generalization. In this work, we study com...
http://arxiv.org/pdf/2301.11459v1
2023-01-26T23:11:03
2023-01-26T23:11:03
{ "title": [ -0.020355225, 0.0052452087, -0.04522705, -0.06958008, -0.024246216, 0.000062286854, -0.10809326, -0.009544373, -0.011131287, 0.03152466, 0.004005432, 0.0099105835, -0.024475098, -0.031234741, -0.043945312, -0.011741638, 0.008804321, ...
http://arxiv.org/abs/2301.11406v1
Beyond Arabic: Software for Perso-Arabic Script Manipulation
Alexander Gutkin, Cibu Johny, Raiomond Doctor, Brian Roark, Richard Sproat
cs.CL, I.2.7; I.7.2; I.7.1
This paper presents an open-source software library that provides a set of finite-state transducer (FST) components and corresponding utilities for manipulating the writing systems of languages that use the Perso-Arabic script. The operations include various levels of script normalization, including visual invariance-p...
http://arxiv.org/pdf/2301.11406v1
2023-01-26T20:37:03
2023-01-26T20:37:03
{ "title": [ -0.038757324, -0.011238098, -0.021530151, -0.041809082, -0.024932861, 0.00667572, -0.026611328, 0.018173218, -0.023361206, 0.037719727, -0.016220093, 0.013061523, 0.024490356, -0.043060303, -0.010826111, -0.0031604767, 0.018753052, 0...
http://arxiv.org/abs/2301.11403v1
Detecting Pump&Dump Stock Market Manipulation from Online Forums
D. Nam, D. B. Skillicorn
cs.SI, cs.CL, cs.LG
The intersection of social media, low-cost trading platforms, and naive investors has created an ideal situation for information-based market manipulations, especially pump&dumps. Manipulators accumulate small-cap stocks, disseminate false information on social media to inflate their price, and sell at the peak. We col...
http://arxiv.org/pdf/2301.11403v1
2023-01-26T20:31:27
2023-01-26T20:31:27
{ "title": [ 0.030197144, 0.024520874, -0.03253174, -0.021530151, 0.030929565, -0.039123535, 0.038970947, -0.025985718, -0.045806885, 0.03390503, -0.025894165, 0.015602112, 0.026107788, -0.024765015, -0.008148193, -0.021575928, -0.015129089, 0.04...
http://arxiv.org/abs/2301.11386v1
Task formulation for Extracting Social Determinants of Health from Clinical Narratives
Manabu Torii, Ian M. Finn, Son Doan, Paul Wang, Elly W. Yang, Daniel S. Zisook
cs.CL, I.2.7
Objective: The 2022 n2c2 NLP Challenge posed identification of social determinants of health (SDOH) in clinical narratives. We present three systems that we developed for the Challenge and discuss the distinctive task formulation used in each of the three systems. Materials and Methods: The first system identifies targ...
http://arxiv.org/pdf/2301.11386v1
2023-01-26T20:00:54
2023-01-26T20:00:54
{ "title": [ -0.003862381, 0.0042686462, -0.0362854, -0.09979248, 0.01184082, -0.0019607544, -0.045532227, 0.011482239, -0.024780273, 0.0019817352, 0.0074386597, 0.027526855, -0.060424805, -0.03930664, 0.0028877258, -0.010025024, -0.015777588, 0....
http://arxiv.org/abs/2301.11367v1
Style-Aware Contrastive Learning for Multi-Style Image Captioning
Yucheng Zhou, Guodong Long
cs.CV, cs.CL
Existing multi-style image captioning methods show promising results in generating a caption with accurate visual content and desired linguistic style. However, existing methods overlook the relationship between linguistic style and visual content. To overcome this drawback, we propose style-aware contrastive learning ...
http://arxiv.org/pdf/2301.11367v1
2023-01-26T19:21:39
2023-01-26T19:21:39
{ "title": [ 0.007820129, 0.011528015, -0.032104492, -0.031951904, 0.008666992, -0.037017822, -0.0073165894, -0.020553589, 0.019256592, 0.05218506, -0.027374268, 0.03491211, 0.0087509155, -0.0039787292, -0.032562256, 0.012641907, -0.007949829, 0....
http://arxiv.org/abs/2301.11362v1
Improving Cross-modal Alignment for Text-Guided Image Inpainting
Yucheng Zhou, Guodong Long
cs.CV, cs.CL
Text-guided image inpainting (TGII) aims to restore missing regions based on a given text in a damaged image. Existing methods are based on a strong vision encoder and a cross-modal fusion model to integrate cross-modal features. However, these methods allocate most of the computation to visual encoding, while light co...
http://arxiv.org/pdf/2301.11362v1
2023-01-26T19:18:27
2023-01-26T19:18:27
{ "title": [ -0.029846191, 0.017242432, -0.033447266, -0.02281189, -0.00080633163, -0.009101868, -0.03302002, -0.03149414, 0.010070801, 0.047332764, -0.03552246, 0.027999878, 0.020355225, -0.017166138, -0.025802612, -0.009674072, 0.0077590942, -0...