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/2301.02807v1
RLAS-BIABC: A Reinforcement Learning-Based Answer Selection Using the BERT Model Boosted by an Improved ABC Algorithm
Hamid Gharagozlou, Javad Mohammadzadeh, Azam Bastanfard, Saeed Shiry Ghidary
cs.CL, cs.AI, cs.IR, I.2.7
Answer selection (AS) is a critical subtask of the open-domain question answering (QA) problem. The present paper proposes a method called RLAS-BIABC for AS, which is established on attention mechanism-based long short-term memory (LSTM) and the bidirectional encoder representations from transformers (BERT) word embedd...
http://arxiv.org/pdf/2301.02807v1
2023-01-07T08:33:05
2023-01-07T08:33:05
{ "title": [ -0.010498047, -0.004928589, -0.00028824806, -0.019226074, -0.018966675, -0.004180908, -0.0071372986, -0.010726929, -0.019821167, 0.018966675, -0.0020542145, 0.017440796, -0.041015625, -0.02520752, -0.013320923, -0.016555786, -0.008499145...
http://arxiv.org/abs/2301.02792v1
Linguistic-style-aware Neural Networks for Fake News Detection
Xinyi Zhou, Jiayu Li, Qinzhou Li, Reza Zafarani
cs.CL, cs.AI
We propose the hierarchical recursive neural network (HERO) to predict fake news by learning its linguistic style, which is distinguishable from the truth, as psychological theories reveal. We first generate the hierarchical linguistic tree of news documents; by doing so, we translate each news document's linguistic st...
http://arxiv.org/pdf/2301.02792v1
2023-01-07T06:48:41
2023-01-07T06:48:41
{ "title": [ -0.017868042, 0.0064811707, -0.03353882, -0.0053253174, 0.001326561, -0.01071167, -0.003227234, -0.02671814, -0.035614014, 0.03604126, 0.0023880005, 0.008590698, 0.010398865, -0.020019531, -0.015419006, 0.0058555603, -0.031707764, 0....
http://arxiv.org/abs/2301.04582v1
A Personalized Utterance Style (PUS) based Dialogue Strategy for Efficient Service Requirement Elicitation
Demin Yu, Min Liu, Zhongjie Wang
cs.CL, cs.AI
With the flourish of services on the Internet, a prerequisite for service providers to precisely deliver services to their customers is to capture user requirements comprehensively, accurately, and efficiently. This is called the ``Service Requirement Elicitation (SRE)'' task. Considering the amount of customers is hug...
http://arxiv.org/pdf/2301.04582v1
2023-01-07T04:10:33
2023-01-07T04:10:33
{ "title": [ 0.02142334, -0.0110321045, -0.0435791, -0.09832764, 0.025878906, -0.014717102, -0.047180176, 0.022277832, -0.009521484, 0.025131226, 0.0032806396, 0.010131836, -0.012992859, -0.030685425, -0.032714844, -0.009864807, -0.025650024, 0.0...
http://arxiv.org/abs/2301.02777v1
Visual Story Generation Based on Emotion and Keywords
Yuetian Chen, Ruohua Li, Bowen Shi, Peiru Liu, Mei Si
cs.AI, cs.CL, cs.LG
Automated visual story generation aims to produce stories with corresponding illustrations that exhibit coherence, progression, and adherence to characters' emotional development. This work proposes a story generation pipeline to co-create visual stories with the users. The pipeline allows the user to control events an...
http://arxiv.org/pdf/2301.02777v1
2023-01-07T03:56:49
2023-01-07T03:56:49
{ "title": [ -0.037017822, -0.01612854, -0.037750244, -0.03125, -0.04071045, 0.007507324, -0.027893066, -0.0027332306, 0.013694763, 0.034118652, -0.019180298, 0.0076408386, -0.008834839, -0.018493652, -0.04434204, -0.002609253, -0.0052604675, 0.0...
http://arxiv.org/abs/2301.02773v1
Building a Parallel Corpus and Training Translation Models Between Luganda and English
Richard Kimera, Daniela N. Rim, Heeyoul Choi
cs.CL
Neural machine translation (NMT) has achieved great successes with large datasets, so NMT is more premised on high-resource languages. This continuously underpins the low resource languages such as Luganda due to the lack of high-quality parallel corpora, so even 'Google translate' does not serve Luganda at the time of...
http://arxiv.org/pdf/2301.02773v1
2023-01-07T03:26:09
2023-01-07T03:26:09
{ "title": [ -0.05001831, -0.0008544922, -0.03265381, -0.0011062622, -0.059051514, 0.0072250366, -0.06896973, -0.034423828, -0.08282471, 0.07287598, 0.011116028, 0.005393982, 0.06378174, -0.047210693, 0.011329651, -0.007843018, 0.010467529, 0.008...
http://arxiv.org/abs/2301.02748v3
Conditional Generation of Paired Antibody Chain Sequences through Encoder-Decoder Language Model
Simon K. S. Chu, Kathy Y. Wei
q-bio.BM, cs.CE, cs.CL
Protein language models (LMs) have been successful in sequence, structural and functional predictions. However, currently, protein LMs are limited to encoder- or decoder-only architectures for single sequences while many biological contexts involve protein-protein interactions. Here, we introduce pAbT5, which models an...
http://arxiv.org/pdf/2301.02748v3
2023-04-04T18:48:22
2023-01-06T23:34:52
{ "title": [ -0.048034668, 0.0022602081, -0.056518555, -0.05734253, -0.0116119385, 0.017425537, -0.068237305, 0.0076065063, 0.023529053, 0.032836914, -0.0028438568, -0.024536133, 0.02067566, -0.04208374, -0.037017822, -0.0017986298, 0.009521484, ...
http://arxiv.org/abs/2301.02724v1
Facilitating Contrastive Learning of Discourse Relational Senses by Exploiting the Hierarchy of Sense Relations
Wanqiu Long, Bonnie Webber
cs.CL, cs.AI
Implicit discourse relation recognition is a challenging task that involves identifying the sense or senses that hold between two adjacent spans of text, in the absence of an explicit connective between them. In both PDTB-2 and PDTB-3, discourse relational senses are organized into a three-level hierarchy ranging from ...
http://arxiv.org/pdf/2301.02724v1
2023-01-06T21:55:48
2023-01-06T21:55:48
{ "title": [ 0.015266418, -0.0044784546, -0.019744873, -0.061035156, -0.008483887, -0.00932312, -0.013183594, -0.012519836, 0.020309448, 0.031280518, -0.01890564, 0.025802612, -0.00056934357, -0.013198853, -0.047912598, 0.0093688965, -0.021194458, ...
http://arxiv.org/abs/2301.02695v1
Witscript 3: A Hybrid AI System for Improvising Jokes in a Conversation
Joe Toplyn
cs.CL, cs.AI, I.2.7; J.5
Previous papers presented Witscript and Witscript 2, AI systems for improvising jokes in a conversation. Witscript generates jokes that rely on wordplay, whereas the jokes generated by Witscript 2 rely on common sense. This paper extends that earlier work by presenting Witscript 3, which generates joke candidates using...
http://arxiv.org/pdf/2301.02695v1
2023-01-06T19:25:46
2023-01-06T19:25:46
{ "title": [ -0.013153076, -0.009635925, -0.0413208, -0.017578125, -0.027175903, -0.026931763, -0.029800415, -0.013755798, 0.02268982, 0.03479004, -0.037017822, -0.00819397, 0.025726318, -0.014099121, -0.048950195, -0.024383545, 0.0060768127, 0.0...
http://arxiv.org/abs/2301.02589v2
Causal Categorization of Mental Health Posts using Transformers
Simranjeet Kaur, Ritika Bhardwaj, Aastha Jain, Muskan Garg, Chandni Saxena
cs.CL
With recent developments in digitization of clinical psychology, NLP research community has revolutionized the field of mental health detection on social media. Existing research in mental health analysis revolves around the cross-sectional studies to classify users' intent on social media. For in-depth analysis, we in...
http://arxiv.org/pdf/2301.02589v2
2023-01-16T03:39:43
2023-01-06T16:37:48
{ "title": [ 0.061065674, 0.013259888, -0.007118225, -0.08154297, 0.0317688, -0.0112838745, -0.013336182, -0.022140503, -0.031982422, 0.013305664, -0.010597229, 0.03375244, -0.0082473755, -0.038482666, 0.0103302, -0.013679504, -0.00036764145, 0.0...
http://arxiv.org/abs/2301.02555v1
"No, to the Right" -- Online Language Corrections for Robotic Manipulation via Shared Autonomy
Yuchen Cui, Siddharth Karamcheti, Raj Palleti, Nidhya Shivakumar, Percy Liang, Dorsa Sadigh
cs.RO, cs.AI, cs.CL, cs.HC, cs.LG
Systems for language-guided human-robot interaction must satisfy two key desiderata for broad adoption: adaptivity and learning efficiency. Unfortunately, existing instruction-following agents cannot adapt, lacking the ability to incorporate online natural language supervision, and even if they could, require hundreds ...
http://arxiv.org/pdf/2301.02555v1
2023-01-06T15:03:27
2023-01-06T15:03:27
{ "title": [ -0.02255249, -0.01777649, -0.01109314, 0.0009636879, -0.0090789795, -0.028076172, 0.019424438, -0.015899658, -0.036010742, 0.037719727, -0.016540527, 0.018722534, 0.0053977966, -0.017822266, -0.019866943, -0.01626587, -0.0138168335, ...
http://arxiv.org/abs/2301.02521v1
SAIDS: A Novel Approach for Sentiment Analysis Informed of Dialect and Sarcasm
Abdelrahman Kaseb, Mona Farouk
cs.CL, cs.AI
Sentiment analysis becomes an essential part of every social network, as it enables decision-makers to know more about users' opinions in almost all life aspects. Despite its importance, there are multiple issues it encounters like the sentiment of the sarcastic text which is one of the main challenges of sentiment ana...
http://arxiv.org/pdf/2301.02521v1
2023-01-06T14:19:46
2023-01-06T14:19:46
{ "title": [ -0.02015686, -0.018508911, -0.01991272, -0.04260254, 0.0090408325, 0.002067566, -0.026672363, -0.006855011, 0.02330017, 0.03604126, -0.02330017, -0.029006958, 0.015052795, -0.006916046, -0.078186035, 0.0049858093, -0.0066337585, 0.04...
http://arxiv.org/abs/2301.02459v1
OPD@NL4Opt: An ensemble approach for the NER task of the optimization problem
Kangxu Wang, Ze Chen, Jiewen Zheng
cs.CL
In this paper, we present an ensemble approach for the NL4Opt competition subtask 1(NER task). For this task, we first fine tune the pretrained language models based on the competition dataset. Then we adopt differential learning rates and adversarial training strategies to enhance the model generalization and robustne...
http://arxiv.org/pdf/2301.02459v1
2023-01-06T11:07:55
2023-01-06T11:07:55
{ "title": [ -0.02078247, -0.0002641678, -0.007144928, -0.035461426, 0.016571045, 0.019515991, -0.010322571, -0.020996094, 0.0010442734, 0.019622803, -0.008918762, 0.031799316, -0.03427124, -0.014282227, -0.026885986, 0.020401001, -0.0069084167, ...
http://arxiv.org/abs/2301.02458v1
Topics as Entity Clusters: Entity-based Topics from Language Models and Graph Neural Networks
Manuel V. Loureiro, Steven Derby, Tri Kurniawan Wijaya
cs.CL, cs.LG
Topic models aim to reveal the latent structure behind a corpus, typically conducted over a bag-of-words representation of documents. In the context of topic modeling, most vocabulary is either irrelevant for uncovering underlying topics or contains strong relationships with relevant concepts, impacting the interpretab...
http://arxiv.org/pdf/2301.02458v1
2023-01-06T10:54:54
2023-01-06T10:54:54
{ "title": [ -0.033203125, 0.00039172173, -0.048919678, -0.043823242, 0.010238647, 0.002708435, -0.03793335, -0.009017944, 0.015136719, 0.024734497, 0.0066184998, 0.029251099, -0.0050468445, -0.01725769, -0.02456665, 0.03286743, -0.008834839, 0.0...
http://arxiv.org/abs/2301.02427v1
Mask-then-Fill: A Flexible and Effective Data Augmentation Framework for Event Extraction
Jun Gao, Changlong Yu, Wei Wang, Huan Zhao, Ruifeng Xu
cs.CL
We present Mask-then-Fill, a flexible and effective data augmentation framework for event extraction. Our approach allows for more flexible manipulation of text and thus can generate more diverse data while keeping the original event structure unchanged as much as possible. Specifically, it first randomly masks out an ...
http://arxiv.org/pdf/2301.02427v1
2023-01-06T09:19:23
2023-01-06T09:19:23
{ "title": [ 0.017196655, 0.020248413, -0.06762695, -0.10321045, 0.049957275, -0.0015926361, -0.05291748, -0.021194458, 0.019348145, 0.017593384, -0.008079529, 0.010368347, -0.00015890598, -0.027862549, -0.032348633, 0.016830444, -0.009735107, 0....
http://arxiv.org/abs/2301.02401v1
You Truly Understand What I Need: Intellectual and Friendly Dialogue Agents grounding Knowledge and Persona
Jungwoo Lim, Myunghoon Kang, Yuna Hur, Seungwon Jung, Jinsung Kim, Yoonna Jang, Dongyub Lee, Hyesung Ji, Donghoon Shin, Seungryong Kim, Heuiseok Lim
cs.CL, cs.AI
To build a conversational agent that interacts fluently with humans, previous studies blend knowledge or personal profile into the pre-trained language model. However, the model that considers knowledge and persona at the same time is still limited, leading to hallucination and a passive way of using personas. We propo...
http://arxiv.org/pdf/2301.02401v1
2023-01-06T06:47:21
2023-01-06T06:47:21
{ "title": [ 0.04638672, -0.0021362305, -0.052978516, -0.045562744, 0.0026817322, -0.05722046, -0.027816772, -0.008232117, 0.012313843, 0.031555176, -0.028213501, -0.018875122, 0.05114746, -0.03515625, -0.019195557, -0.023254395, -0.0124435425, 0...
http://arxiv.org/abs/2301.03373v1
Chatbots As Fluent Polyglots: Revisiting Breakthrough Code Snippets
David Noever, Kevin Williams
cs.LG, cs.CL, cs.SE
The research applies AI-driven code assistants to analyze a selection of influential computer code that has shaped modern technology, including email, internet browsing, robotics, and malicious software. The original contribution of this study was to examine half of the most significant code advances in the last 50 yea...
http://arxiv.org/pdf/2301.03373v1
2023-01-05T23:17:17
2023-01-05T23:17:17
{ "title": [ -0.022644043, -0.025177002, -0.025421143, -0.04711914, -0.022079468, -0.0024223328, -0.019363403, 0.013061523, -0.0077285767, 0.036621094, -0.0119018555, 0.009010315, -0.0011968613, -0.01927185, -0.03189087, -0.0028629303, -0.004016876, ...
http://arxiv.org/abs/2301.02299v1
Sequentially Controlled Text Generation
Alexander Spangher, Xinyu Hua, Yao Ming, Nanyun Peng
cs.CL, cs.AI, cs.LG
While GPT-2 generates sentences that are remarkably human-like, longer documents can ramble and do not follow human-like writing structure. We study the problem of imposing structure on long-range text. We propose a novel controlled text generation task, sequentially controlled text generation, and identify a dataset, ...
http://arxiv.org/pdf/2301.02299v1
2023-01-05T21:23:51
2023-01-05T21:23:51
{ "title": [ -0.017364502, -0.005947113, -0.02798462, -0.06530762, -0.012313843, 0.007259369, -0.05657959, -0.018249512, -0.006965637, 0.05331421, -0.0011806488, 0.007282257, -0.0041999817, -0.024642944, -0.043884277, 0.016815186, -0.004852295, 0...
http://arxiv.org/abs/2301.02284v1
Unsupervised Broadcast News Summarization; a comparative study on Maximal Marginal Relevance (MMR) and Latent Semantic Analysis (LSA)
Majid Ramezani, Mohammad-Salar Shahryari, Amir-Reza Feizi-Derakhshi, Mohammad-Reza Feizi-Derakhshi
cs.CL
The methods of automatic speech summarization are classified into two groups: supervised and unsupervised methods. Supervised methods are based on a set of features, while unsupervised methods perform summarization based on a set of rules. Latent Semantic Analysis (LSA) and Maximal Marginal Relevance (MMR) are consider...
http://arxiv.org/pdf/2301.02284v1
2023-01-05T20:13:44
2023-01-05T20:13:44
{ "title": [ -0.01234436, -0.008323669, -0.04046631, -0.03942871, -0.026428223, -0.015686035, -0.018188477, 0.0076942444, -0.051330566, 0.07495117, -0.018417358, 0.042510986, 0.013908386, -0.0058784485, -0.013916016, -0.010421753, 0.02444458, -0....
http://arxiv.org/abs/2301.02275v2
Language as a Latent Sequence: deep latent variable models for semi-supervised paraphrase generation
Jialin Yu, Alexandra I. Cristea, Anoushka Harit, Zhongtian Sun, Olanrewaju Tahir Aduragba, Lei Shi, Noura Al Moubayed
cs.CL, cs.LG
This paper explores deep latent variable models for semi-supervised paraphrase generation, where the missing target pair for unlabelled data is modelled as a latent paraphrase sequence. We present a novel unsupervised model named variational sequence auto-encoding reconstruction (VSAR), which performs latent sequence i...
http://arxiv.org/pdf/2301.02275v2
2023-09-08T21:03:14
2023-01-05T19:35:30
{ "title": [ -0.023803711, -0.008934021, -0.020980835, -0.07220459, -0.0262146, 0.014556885, -0.06939697, 0.015129089, 0.0046920776, 0.02571106, -0.02104187, 0.01663208, 0.018188477, -0.014846802, -0.018936157, 0.0071983337, -0.022277832, 0.02435...
http://arxiv.org/abs/2301.02241v1
CiT: Curation in Training for Effective Vision-Language Data
Hu Xu, Saining Xie, Po-Yao Huang, Licheng Yu, Russell Howes, Gargi Ghosh, Luke Zettlemoyer, Christoph Feichtenhofer
cs.CV, cs.CL
Large vision-language models are generally applicable to many downstream tasks, but come at an exorbitant training cost that only large institutions can afford. This paper trades generality for efficiency and presents Curation in Training (CiT), a simple and efficient vision-text learning algorithm that couples a data ...
http://arxiv.org/pdf/2301.02241v1
2023-01-05T18:59:57
2023-01-05T18:59:57
{ "title": [ -0.007926941, -0.024887085, -0.05090332, -0.048919678, -0.016052246, -0.034606934, -0.016845703, 0.005504608, 0.0013217926, 0.027618408, -0.038024902, -0.0015707016, 0.010147095, -0.025665283, -0.027526855, -0.019866943, 0.007381439, ...
http://arxiv.org/abs/2301.02228v3
MedKLIP: Medical Knowledge Enhanced Language-Image Pre-Training in Radiology
Chaoyi Wu, Xiaoman Zhang, Ya Zhang, Yanfeng Wang, Weidi Xie
eess.IV, cs.CL, cs.CV
In this paper, we consider enhancing medical visual-language pre-training (VLP) with domain-specific knowledge, by exploiting the paired image-text reports from the radiological daily practice. In particular, we make the following contributions: First, unlike existing works that directly process the raw reports, we ado...
http://arxiv.org/pdf/2301.02228v3
2023-04-03T09:57:51
2023-01-05T18:55:09
{ "title": [ 0.0006775856, -0.016143799, -0.036987305, -0.03845215, -0.019302368, -0.01651001, -0.010147095, -0.0078086853, -0.02720642, 0.023544312, 0.004524231, 0.017364502, 0.0066108704, -0.05380249, 0.011276245, -0.004901886, 0.020690918, -0....
http://arxiv.org/abs/2302.10291v1
Can Large Language Models Change User Preference Adversarially?
Varshini Subhash
cs.CL, cs.LG
Pretrained large language models (LLMs) are becoming increasingly powerful and ubiquitous in mainstream applications such as being a personal assistant, a dialogue model, etc. As these models become proficient in deducing user preferences and offering tailored assistance, there is an increasing concern about the abilit...
http://arxiv.org/pdf/2302.10291v1
2023-01-05T18:49:21
2023-01-05T18:49:21
{ "title": [ -0.0043411255, 0.010726929, -0.024108887, -0.014976501, -0.0088272095, -0.013969421, -0.010955811, -0.0206604, -0.011131287, 0.031402588, -0.022613525, -0.016586304, -0.008201599, -0.01776123, -0.03353882, -0.015419006, -0.017410278, ...
http://arxiv.org/abs/2301.02195v1
Towards Autoformalization of Mathematics and Code Correctness: Experiments with Elementary Proofs
Garett Cunningham, Razvan C. Bunescu, David Juedes
cs.CL, cs.AI, cs.LO
The ever-growing complexity of mathematical proofs makes their manual verification by mathematicians very cognitively demanding. Autoformalization seeks to address this by translating proofs written in natural language into a formal representation that is computer-verifiable via interactive theorem provers. In this pap...
http://arxiv.org/pdf/2301.02195v1
2023-01-05T17:56:00
2023-01-05T17:56:00
{ "title": [ -0.0051460266, 0.021148682, -0.04156494, -0.072387695, 0.031921387, -0.009849548, -0.036346436, -0.0014791489, -0.014755249, 0.03439331, -0.0033626556, 0.009216309, 0.004432678, -0.021392822, -0.023284912, -0.0015239716, -0.011962891, ...
http://arxiv.org/abs/2301.02120v1
Reprogramming Pretrained Language Models for Protein Sequence Representation Learning
Ria Vinod, Pin-Yu Chen, Payel Das
cs.LG, cs.CL, q-bio.BM
Machine Learning-guided solutions for protein learning tasks have made significant headway in recent years. However, success in scientific discovery tasks is limited by the accessibility of well-defined and labeled in-domain data. To tackle the low-data constraint, recent adaptions of deep learning models pretrained on...
http://arxiv.org/pdf/2301.02120v1
2023-01-05T15:55:18
2023-01-05T15:55:18
{ "title": [ -0.030883789, -0.01737976, -0.04272461, -0.055847168, -0.014984131, 0.0002399683, -0.04824829, 0.021957397, 0.00868988, 0.03479004, -0.014892578, -0.003446579, 0.011405945, -0.025253296, -0.03137207, 0.014877319, 0.0082473755, 0.0342...
http://arxiv.org/abs/2301.02113v1
Anaphora Resolution in Dialogue: System Description (CODI-CRAC 2022 Shared Task)
Tatiana Anikina, Natalia Skachkova, Joseph Renner, Priyansh Trivedi
cs.CL, cs.LG
We describe three models submitted for the CODI-CRAC 2022 shared task. To perform identity anaphora resolution, we test several combinations of the incremental clustering approach based on the Workspace Coreference System (WCS) with other coreference models. The best result is achieved by adding the ''cluster merging''...
http://arxiv.org/pdf/2301.02113v1
2023-01-05T15:42:17
2023-01-05T15:42:17
{ "title": [ -0.009536743, -0.0025806427, -0.040771484, -0.063964844, 0.0036067963, -0.010673523, -0.061279297, -0.0018815994, -0.034729004, 0.052764893, 0.0041389465, 0.007965088, 0.029891968, -0.040863037, -0.0072631836, 0.01449585, -0.013458252, ...
http://arxiv.org/abs/2301.02111v1
Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers
Chengyi Wang, Sanyuan Chen, Yu Wu, Ziqiang Zhang, Long Zhou, Shujie Liu, Zhuo Chen, Yanqing Liu, Huaming Wang, Jinyu Li, Lei He, Sheng Zhao, Furu Wei
cs.CL, cs.SD, eess.AS
We introduce a language modeling approach for text to speech synthesis (TTS). Specifically, we train a neural codec language model (called Vall-E) using discrete codes derived from an off-the-shelf neural audio codec model, and regard TTS as a conditional language modeling task rather than continuous signal regression ...
http://arxiv.org/pdf/2301.02111v1
2023-01-05T15:37:15
2023-01-05T15:37:15
{ "title": [ -0.00983429, -0.02406311, -0.0059814453, -0.0019664764, -0.07183838, -0.011360168, -0.036102295, 0.014533997, 0.014328003, 0.057556152, -0.020935059, -0.018585205, 0.05407715, -0.028915405, -0.025527954, 0.020019531, 0.0023078918, 0....
http://arxiv.org/abs/2301.02071v1
Towards Table-to-Text Generation with Pretrained Language Model: A Table Structure Understanding and Text Deliberating Approach
Miao Chen, Xinjiang Lu, Tong Xu, Yanyan Li, Jingbo Zhou, Dejing Dou, Hui Xiong
cs.CL, cs.AI
Although remarkable progress on the neural table-to-text methods has been made, the generalization issues hinder the applicability of these models due to the limited source tables. Large-scale pretrained language models sound like a promising solution to tackle such issues. However, how to effectively bridge the gap be...
http://arxiv.org/pdf/2301.02071v1
2023-01-05T14:03:26
2023-01-05T14:03:26
{ "title": [ -0.02494812, -0.009559631, -0.03869629, -0.04837036, -0.01210022, 0.004611969, -0.075805664, -0.013626099, 0.0053710938, 0.04498291, -0.0059928894, 0.017242432, 0.016540527, -0.0010995865, -0.059143066, 0.003698349, -0.009010315, 0.0...
http://arxiv.org/abs/2301.02057v3
TextDescriptives: A Python package for calculating a large variety of metrics from text
Lasse Hansen, Ludvig Renbo Olsen, Kenneth Enevoldsen
cs.CL
TextDescriptives is a Python package for calculating a large variety of metrics from text. It is built on top of spaCy and can be easily integrated into existing workflows. The package has already been used for analysing the linguistic stability of clinical texts, creating features for predicting neuropsychiatric condi...
http://arxiv.org/pdf/2301.02057v3
2023-03-28T14:57:59
2023-01-05T13:19:17
{ "title": [ -0.043426514, -0.023757935, -0.020401001, -0.030639648, -0.025314331, -0.009056091, -0.028213501, -0.022338867, 0.016494751, 0.029418945, -0.023513794, 0.007369995, 0.024902344, -0.00868988, -0.029724121, 0.0012464523, 0.0119018555, ...
http://arxiv.org/abs/2301.02010v1
HIT-SCIR at MMNLU-22: Consistency Regularization for Multilingual Spoken Language Understanding
Bo Zheng, Zhouyang Li, Fuxuan Wei, Qiguang Chen, Libo Qin, Wanxiang Che
cs.CL
Multilingual spoken language understanding (SLU) consists of two sub-tasks, namely intent detection and slot filling. To improve the performance of these two sub-tasks, we propose to use consistency regularization based on a hybrid data augmentation strategy. The consistency regularization enforces the predicted distri...
http://arxiv.org/pdf/2301.02010v1
2023-01-05T11:21:15
2023-01-05T11:21:15
{ "title": [ -0.024047852, -0.011245728, -0.04562378, -0.080444336, -0.0026626587, 0.019638062, -0.08068848, 0.027252197, -0.0047416687, 0.0317688, -0.017501831, -0.008407593, 0.0066223145, -0.01914978, -0.04135132, -0.007774353, -0.019989014, 0....
http://arxiv.org/abs/2301.01982v2
Emotion-Cause Pair Extraction as Question Answering
Huu-Hiep Nguyen, Minh-Tien Nguyen
cs.CL
The task of Emotion-Cause Pair Extraction (ECPE) aims to extract all potential emotion-cause pairs of a document without any annotation of emotion or cause clauses. Previous approaches on ECPE have tried to improve conventional two-step processing schemes by using complex architectures for modeling emotion-cause intera...
http://arxiv.org/pdf/2301.01982v2
2023-01-06T01:48:34
2023-01-05T09:33:41
{ "title": [ 0.001912117, -0.0011301041, -0.028793335, -0.06121826, 0.004917145, -0.0105896, -0.02645874, -0.00016713142, 0.04699707, 0.0014896393, -0.004650116, -0.015960693, -0.004787445, -0.030899048, -0.04663086, 0.0047035217, 0.0010299683, 0...
http://arxiv.org/abs/2301.01967v1
A Survey of Code-switching: Linguistic and Social Perspectives for Language Technologies
A. Seza Doğruöz, Sunayana Sitaram, Barbara E. Bullock, Almeida Jacqueline Toribio
cs.CL
The analysis of data in which multiple languages are represented has gained popularity among computational linguists in recent years. So far, much of this research focuses mainly on the improvement of computational methods and largely ignores linguistic and social aspects of C-S discussed across a wide range of languag...
http://arxiv.org/pdf/2301.01967v1
2023-01-05T09:08:04
2023-01-05T09:08:04
{ "title": [ -0.029586792, -0.005306244, -0.02734375, -0.015075684, 0.008201599, -0.0064468384, 0.0017929077, -0.0016565323, -0.015686035, 0.0395813, -0.0023536682, 0.007904053, 0.0074882507, -0.030319214, -0.031188965, 0.009307861, -0.02331543, ...
http://arxiv.org/abs/2301.01949v1
SPRING: Situated Conversation Agent Pretrained with Multimodal Questions from Incremental Layout Graph
Yuxing Long, Binyuan Hui, Fulong Ye, Yanyang Li, Zhuoxin Han, Caixia Yuan, Yongbin Li, Xiaojie Wang
cs.CL, cs.AI, cs.CV, cs.LG, cs.MM
Existing multimodal conversation agents have shown impressive abilities to locate absolute positions or retrieve attributes in simple scenarios, but they fail to perform well when complex relative positions and information alignments are involved, which poses a bottleneck in response quality. In this paper, we propose ...
http://arxiv.org/pdf/2301.01949v1
2023-01-05T08:03:47
2023-01-05T08:03:47
{ "title": [ -0.0149002075, -0.01285553, -0.056152344, -0.06530762, -0.02468872, -0.012741089, -0.046295166, 0.024139404, 0.008857727, 0.006389618, -0.026550293, 0.015670776, -0.016326904, -0.02583313, -0.0055160522, -0.01991272, -0.014503479, 0....
http://arxiv.org/abs/2301.01768v1
The political ideology of conversational AI: Converging evidence on ChatGPT's pro-environmental, left-libertarian orientation
Jochen Hartmann, Jasper Schwenzow, Maximilian Witte
cs.CL, cs.CY
Conversational artificial intelligence (AI) disrupts how humans interact with technology. Recently, OpenAI introduced ChatGPT, a state-of-the-art dialogue model that can converse with its human counterparts with unprecedented capabilities. ChatGPT has witnessed tremendous attention from the media, academia, industry, a...
http://arxiv.org/pdf/2301.01768v1
2023-01-05T07:13:13
2023-01-05T07:13:13
{ "title": [ 0.009918213, -0.0037117004, -0.049987793, -0.026931763, -0.0018692017, -0.040496826, 0.005744934, -0.0017604828, -0.026855469, 0.032073975, -0.017364502, 0.029174805, -0.028274536, -0.019165039, -0.020187378, -0.02178955, -0.04324341, ...
http://arxiv.org/abs/2301.01935v1
Topic Segmentation Model Focusing on Local Context
Jeonghwan Lee, Jiyeong Han, Sunghoon Baek, Min Song
cs.CL, cs.AI
Topic segmentation is important in understanding scientific documents since it can not only provide better readability but also facilitate downstream tasks such as information retrieval and question answering by creating appropriate sections or paragraphs. In the topic segmentation task, topic coherence is critical in ...
http://arxiv.org/pdf/2301.01935v1
2023-01-05T06:57:42
2023-01-05T06:57:42
{ "title": [ -0.010528564, 0.024673462, -0.03829956, -0.033691406, -0.00066041946, -0.013038635, -0.0209198, -0.03503418, 0.013671875, 0.027359009, -0.012252808, 0.01977539, -0.0050697327, -0.03250122, -0.00025844574, 0.014289856, 0.00831604, 0.0...
http://arxiv.org/abs/2301.01893v1
GIVL: Improving Geographical Inclusivity of Vision-Language Models with Pre-Training Methods
Da Yin, Feng Gao, Govind Thattai, Michael Johnston, Kai-Wei Chang
cs.CV, cs.AI, cs.CL
A key goal for the advancement of AI is to develop technologies that serve the needs not just of one group but of all communities regardless of their geographical region. In fact, a significant proportion of knowledge is locally shared by people from certain regions but may not apply equally in other regions because of...
http://arxiv.org/pdf/2301.01893v1
2023-01-05T03:43:45
2023-01-05T03:43:45
{ "title": [ -0.0058059692, -0.014205933, -0.04159546, -0.059448242, 0.0019359589, -0.016616821, -0.043914795, -0.005332947, 0.02027893, 0.036193848, -0.03137207, -0.0030326843, 0.029525757, -0.0128479, -0.039886475, -0.001156807, 0.004901886, 0....
http://arxiv.org/abs/2301.01874v1
Critical Perspectives: A Benchmark Revealing Pitfalls in PerspectiveAPI
Lorena Piedras, Lucas Rosenblatt, Julia Wilkins
cs.CL, cs.CY, cs.LG, I.2.7
Detecting "toxic" language in internet content is a pressing social and technical challenge. In this work, we focus on PERSPECTIVE from Jigsaw, a state-of-the-art tool that promises to score the "toxicity" of text, with a recent model update that claims impressive results (Lees et al., 2022). We seek to challenge certa...
http://arxiv.org/pdf/2301.01874v1
2023-01-05T02:12:47
2023-01-05T02:12:47
{ "title": [ -0.028411865, -0.004131317, -0.004360199, -0.006298065, -0.037994385, -0.0262146, -0.018051147, -0.041870117, -0.01159668, 0.02218628, -0.025115967, -0.02848816, -0.033569336, -0.051452637, -0.045776367, -0.023452759, -0.0026550293, ...
http://arxiv.org/abs/2301.01821v1
Parameter-Efficient Fine-Tuning Design Spaces
Jiaao Chen, Aston Zhang, Xingjian Shi, Mu Li, Alex Smola, Diyi Yang
cs.CL, cs.AI
Parameter-efficient fine-tuning aims to achieve performance comparable to fine-tuning, using fewer trainable parameters. Several strategies (e.g., Adapters, prefix tuning, BitFit, and LoRA) have been proposed. However, their designs are hand-crafted separately, and it remains unclear whether certain design patterns exi...
http://arxiv.org/pdf/2301.01821v1
2023-01-04T21:00:18
2023-01-04T21:00:18
{ "title": [ -0.023544312, -0.017654419, -0.013809204, -0.058685303, -0.0038261414, 0.0067863464, -0.027145386, 0.0028591156, -0.018127441, 0.033843994, -0.010948181, 0.011787415, -0.013336182, -0.017944336, -0.014801025, 0.016082764, 0.007850647, ...
http://arxiv.org/abs/2301.01808v1
MessageNet: Message Classification using Natural Language Processing and Meta-data
Adar Kahana, Oren Elisha
cs.LG, cs.CL
In this paper we propose a new Deep Learning (DL) approach for message classification. Our method is based on the state-of-the-art Natural Language Processing (NLP) building blocks, combined with a novel technique for infusing the meta-data input that is typically available in messages such as the sender information, t...
http://arxiv.org/pdf/2301.01808v1
2023-01-04T20:11:00
2023-01-04T20:11:00
{ "title": [ -0.028915405, -0.01486969, -0.040924072, -0.04559326, 0.018615723, 0.0076675415, -0.031036377, -0.031555176, 0.01322937, 0.032470703, -0.010063171, -0.008049011, 0.00023162365, -0.021881104, -0.066223145, -0.00024676323, -0.017425537, ...
http://arxiv.org/abs/2301.03403v4
A comprehensive review of automatic text summarization techniques: method, data, evaluation and coding
Daniel O. Cajueiro, Arthur G. Nery, Igor Tavares, Maísa K. De Melo, Silvia A. dos Reis, Li Weigang, Victor R. R. Celestino
cs.CL, cs.LG
We provide a literature review about Automatic Text Summarization (ATS) systems. We consider a citation-based approach. We start with some popular and well-known papers that we have in hand about each topic we want to cover and we have tracked the "backward citations" (papers that are cited by the set of papers we knew...
http://arxiv.org/pdf/2301.03403v4
2023-10-04T00:46:08
2023-01-04T19:20:18
{ "title": [ -0.025100708, -0.0003452301, -0.022140503, -0.05593872, 0.004737854, 0.010932922, -0.031402588, -0.020812988, -0.0039863586, 0.053894043, -0.0074501038, 0.034179688, 0.008125305, 0.002620697, -0.05731201, 0.024719238, -0.0031375885, ...
http://arxiv.org/abs/2301.01764v2
UniHD at TSAR-2022 Shared Task: Is Compute All We Need for Lexical Simplification?
Dennis Aumiller, Michael Gertz
cs.CL
Previous state-of-the-art models for lexical simplification consist of complex pipelines with several components, each of which requires deep technical knowledge and fine-tuned interaction to achieve its full potential. As an alternative, we describe a frustratingly simple pipeline based on prompted GPT-3 responses, be...
http://arxiv.org/pdf/2301.01764v2
2023-01-05T15:22:05
2023-01-04T18:59:20
{ "title": [ -0.033111572, -0.01576233, -0.040161133, -0.073791504, -0.01725769, 0.012313843, -0.0736084, 0.014335632, 0.00472641, 0.047088623, -0.009170532, 0.007358551, 0.02609253, -0.025054932, -0.014907837, 0.024765015, 0.0013561249, 0.020355...
http://arxiv.org/abs/2301.01751v2
Iterated Decomposition: Improving Science Q&A by Supervising Reasoning Processes
Justin Reppert, Ben Rachbach, Charlie George, Luke Stebbing, Jungwon Byun, Maggie Appleton, Andreas Stuhlmüller
cs.CL, cs.AI, cs.HC
Language models (LMs) can perform complex reasoning either end-to-end, with hidden latent state, or compositionally, with transparent intermediate state. Composition offers benefits for interpretability and safety, but may need workflow support and infrastructure to remain competitive. We describe iterated decompositio...
http://arxiv.org/pdf/2301.01751v2
2023-01-05T01:35:43
2023-01-04T18:34:25
{ "title": [ -0.009315491, -0.014656067, -0.04055786, -0.06097412, 0.005554199, -0.009727478, -0.018951416, 0.0065689087, -0.006996155, -0.0036087036, -0.00699234, 0.014343262, -0.018249512, -0.0345459, -0.0043525696, -0.0050086975, -0.012832642, ...
http://arxiv.org/abs/2301.01664v2
Multi-Aspect Explainable Inductive Relation Prediction by Sentence Transformer
Zhixiang Su, Di Wang, Chunyan Miao, Lizhen Cui
cs.CL, cs.LG
Recent studies on knowledge graphs (KGs) show that path-based methods empowered by pre-trained language models perform well in the provision of inductive and explainable relation predictions. In this paper, we introduce the concepts of relation path coverage and relation path confidence to filter out unreliable paths p...
http://arxiv.org/pdf/2301.01664v2
2023-05-01T09:52:04
2023-01-04T15:33:49
{ "title": [ 0.0146865845, -0.0022678375, -0.0041236877, -0.08605957, -0.011375427, 0.013832092, -0.041015625, 0.010826111, -0.005180359, 0.01486969, -0.002696991, 0.003446579, 0.01260376, -0.0236969, -0.022323608, -0.001499176, -0.025375366, 0.0...
http://arxiv.org/abs/2301.01621v3
Grammar construction methods for extended deterministic expressions
Xiaoying Mou, Haiming Chen
cs.CL
Extended regular expressions with counting and interleaving are widely used in practice. However the related theoretical studies for this kind of expressions currently cannot meet the need of practical work. This paper develops syntax definitions for extended deterministic expressions and their subclasses, hope to comp...
http://arxiv.org/pdf/2301.01621v3
2023-02-02T05:22:04
2023-01-04T13:49:14
{ "title": [ -0.013412476, -0.006378174, 0.0007171631, -0.08709717, -0.008598328, 0.019500732, -0.053497314, 0.0062828064, -0.0093688965, 0.00844574, 0.005558014, -0.007709503, -0.035583496, -0.043518066, -0.014595032, 0.005622864, -0.0234375, 0....
http://arxiv.org/abs/2302.05289v1
Rumor Classification through a Multimodal Fusion Framework and Ensemble Learning
Abderrazek Azri, Cécile Favre, Nouria Harbi, Jérôme Darmont, Camille Noûs
cs.CV, cs.AI, cs.CL, cs.SI
The proliferation of rumors on social media has become a major concern due to its ability to create a devastating impact. Manually assessing the veracity of social media messages is a very time-consuming task that can be much helped by machine learning. Most message veracity verification methods only exploit textual co...
http://arxiv.org/pdf/2302.05289v1
2023-01-04T13:15:08
2023-01-04T13:15:08
{ "title": [ -0.016204834, 0.012458801, -0.05911255, -0.030654907, -0.008224487, -0.028427124, -0.035247803, -0.017318726, 0.011375427, 0.034179688, -0.0362854, -0.010688782, -0.007881165, -0.006580353, -0.060699463, -0.024169922, -0.030258179, 0...
http://arxiv.org/abs/2301.01579v1
Learning Ambiguity from Crowd Sequential Annotations
Xiaolei Lu
cs.CL, cs.AI, cs.HC, cs.LG
Most crowdsourcing learning methods treat disagreement between annotators as noisy labelings while inter-disagreement among experts is often a good indicator for the ambiguity and uncertainty that is inherent in natural language. In this paper, we propose a framework called Learning Ambiguity from Crowd Sequential Anno...
http://arxiv.org/pdf/2301.01579v1
2023-01-04T12:53:56
2023-01-04T12:53:56
{ "title": [ -0.004875183, 0.022338867, -0.039916992, -0.0713501, 0.011177063, -0.0064964294, -0.034606934, -0.002128601, 0.0129852295, 0.031585693, -0.034973145, -0.012687683, -0.023132324, -0.008720398, -0.07220459, -0.020217896, 0.014793396, 0...
http://arxiv.org/abs/2301.01456v1
Audio-Visual Efficient Conformer for Robust Speech Recognition
Maxime Burchi, Radu Timofte
cs.CV, cs.CL, cs.SD, eess.AS
End-to-end Automatic Speech Recognition (ASR) systems based on neural networks have seen large improvements in recent years. The availability of large scale hand-labeled datasets and sufficient computing resources made it possible to train powerful deep neural networks, reaching very low Word Error Rate (WER) on academ...
http://arxiv.org/pdf/2301.01456v1
2023-01-04T05:36:56
2023-01-04T05:36:56
{ "title": [ -0.00016498566, 0.012809753, -0.055755615, -0.09539795, 0.016448975, 0.00041937828, -0.056030273, 0.010971069, 0.0042648315, 0.05090332, -0.03894043, 0.020477295, 0.009239197, -0.004360199, -0.056793213, -0.010154724, -0.006122589, 0...
http://arxiv.org/abs/2301.01378v1
An ensemble-based framework for mispronunciation detection of Arabic phonemes
Sukru Selim Calik, Ayhan Kucukmanisa, Zeynep Hilal Kilimci
cs.SD, cs.CL, cs.LG, eess.AS
Determination of mispronunciations and ensuring feedback to users are maintained by computer-assisted language learning (CALL) systems. In this work, we introduce an ensemble model that defines the mispronunciation of Arabic phonemes and assists learning of Arabic, effectively. To the best of our knowledge, this is the...
http://arxiv.org/pdf/2301.01378v1
2023-01-03T22:17:08
2023-01-03T22:17:08
{ "title": [ 0.004722595, -0.0034637451, -0.04611206, -0.05807495, 0.0076446533, -0.015640259, -0.011367798, 0.014442444, 0.0044898987, 0.033569336, -0.03366089, -0.0006213188, 0.0004339218, -0.036010742, -0.035125732, -0.011268616, -0.019866943, ...
http://arxiv.org/abs/2301.01269v1
Average Is Not Enough: Caveats of Multilingual Evaluation
Matúš Pikuliak, Marián Šimko
cs.CL
This position paper discusses the problem of multilingual evaluation. Using simple statistics, such as average language performance, might inject linguistic biases in favor of dominant language families into evaluation methodology. We argue that a qualitative analysis informed by comparative linguistics is needed for m...
http://arxiv.org/pdf/2301.01269v1
2023-01-03T18:23:42
2023-01-03T18:23:42
{ "title": [ -0.026550293, 0.006374359, -0.020553589, -0.028335571, 0.0043640137, -0.0037174225, -0.018997192, -0.027648926, -0.029663086, 0.023254395, -0.0017938614, 0.0012569427, -0.03692627, -0.032348633, -0.040283203, -0.009056091, -0.032226562, ...
http://arxiv.org/abs/2301.01181v7
Large Language Models as Corporate Lobbyists
John J. Nay
cs.CL, cs.CY
We demonstrate a proof-of-concept of a large language model conducting corporate lobbying related activities. An autoregressive large language model (OpenAI's text-davinci-003) determines if proposed U.S. Congressional bills are relevant to specific public companies and provides explanations and confidence levels. For ...
http://arxiv.org/pdf/2301.01181v7
2023-01-28T20:49:33
2023-01-03T16:25:52
{ "title": [ -0.0074005127, -0.019592285, -0.03152466, -0.052764893, -0.06628418, -0.0087890625, -0.04345703, 0.031188965, -0.057250977, 0.031677246, -0.05215454, 0.022476196, -0.008491516, -0.026046753, 0.022903442, -0.037261963, 0.0016345978, 0...
http://arxiv.org/abs/2301.01172v1
A Survey On Few-shot Knowledge Graph Completion with Structural and Commonsense Knowledge
Haodi Ma, Daisy Zhe Wang
cs.CL, cs.AI, cs.LG
Knowledge graphs (KG) have served as the key component of various natural language processing applications. Commonsense knowledge graphs (CKG) are a special type of KG, where entities and relations are composed of free-form text. However, previous works in KG completion and CKG completion suffer from long-tail relation...
http://arxiv.org/pdf/2301.01172v1
2023-01-03T16:00:09
2023-01-03T16:00:09
{ "title": [ -0.016784668, -0.03930664, -0.042144775, -0.07904053, -0.027557373, 0.021453857, -0.06549072, 0.027557373, 0.020050049, 0.00919342, -0.029953003, 0.023956299, -0.02658081, -0.009292603, -0.031143188, -0.006793976, 0.009590149, 0.0041...
http://arxiv.org/abs/2301.01162v1
Language Models are Drummers: Drum Composition with Natural Language Pre-Training
Li Zhang, Chris Callison-Burch
cs.SD, cs.CL, eess.AS
Automatic music generation with artificial intelligence typically requires a large amount of data which is hard to obtain for many less common genres and musical instruments. To tackle this issue, we present ongoing work and preliminary findings on the possibility for deep models to transfer knowledge from language to ...
http://arxiv.org/pdf/2301.01162v1
2023-01-03T15:47:53
2023-01-03T15:47:53
{ "title": [ -0.01687622, -0.01789856, -0.020767212, -0.054473877, -0.046203613, -0.003232956, -0.07745361, 0.0006160736, -0.010597229, 0.016540527, -0.041381836, 0.013763428, 0.012886047, -0.029800415, -0.0031375885, -0.025848389, 0.0066108704, ...
http://arxiv.org/abs/2301.01067v1
Towards Knowledge-Intensive Text-to-SQL Semantic Parsing with Formulaic Knowledge
Longxu Dou, Yan Gao, Xuqi Liu, Mingyang Pan, Dingzirui Wang, Wanxiang Che, Dechen Zhan, Min-Yen Kan, Jian-Guang Lou
cs.CL
In this paper, we study the problem of knowledge-intensive text-to-SQL, in which domain knowledge is necessary to parse expert questions into SQL queries over domain-specific tables. We formalize this scenario by building a new Chinese benchmark KnowSQL consisting of domain-specific questions covering various domains. ...
http://arxiv.org/pdf/2301.01067v1
2023-01-03T12:37:47
2023-01-03T12:37:47
{ "title": [ -0.02482605, -0.0012979507, -0.06829834, -0.09918213, 0.020004272, 0.0044670105, -0.06628418, 0.038360596, 0.008926392, 0.039733887, -0.049682617, 0.01638794, -0.00038528442, 0.0016431808, -0.0647583, -0.016357422, -0.005458832, 0.03...
http://arxiv.org/abs/2301.01064v1
PIE-QG: Paraphrased Information Extraction for Unsupervised Question Generation from Small Corpora
Dinesh Nagumothu, Bahadorreza Ofoghi, Guangyan Huang, Peter W. Eklund
cs.CL, cs.AI
Supervised Question Answering systems (QA systems) rely on domain-specific human-labeled data for training. Unsupervised QA systems generate their own question-answer training pairs, typically using secondary knowledge sources to achieve this outcome. Our approach (called PIE-QG) uses Open Information Extraction (OpenI...
http://arxiv.org/pdf/2301.01064v1
2023-01-03T12:20:51
2023-01-03T12:20:51
{ "title": [ -0.021453857, -0.007347107, -0.02178955, -0.04586792, -0.014984131, 0.013839722, -0.037475586, -0.023513794, 0.01914978, 0.0178833, -0.03201294, 0.005050659, -0.012123108, -0.009727478, -0.047302246, -0.00030207634, -0.012458801, 0.0...
http://arxiv.org/abs/2301.01020v1
Supervised Acoustic Embeddings And Their Transferability Across Languages
Sreepratha Ram, Hanan Aldarmaki
cs.CL, cs.SD, eess.AS
In speech recognition, it is essential to model the phonetic content of the input signal while discarding irrelevant factors such as speaker variations and noise, which is challenging in low-resource settings. Self-supervised pre-training has been proposed as a way to improve both supervised and unsupervised speech rec...
http://arxiv.org/pdf/2301.01020v1
2023-01-03T09:37:24
2023-01-03T09:37:24
{ "title": [ -0.005470276, -0.0099105835, -0.015731812, -0.07965088, -0.006099701, 0.012825012, -0.06933594, 0.016296387, 0.0073432922, 0.012901306, -0.017150879, 0.0005698204, -0.0028858185, -0.027450562, -0.016174316, 0.009132385, -0.016708374, ...
http://arxiv.org/abs/2301.01015v4
Semi-Structured Object Sequence Encoders
Rudra Murthy V, Riyaz Bhat, Chulaka Gunasekara, Siva Sankalp Patel, Hui Wan, Tejas Indulal Dhamecha, Danish Contractor, Marina Danilevsky
cs.CV, cs.AI, cs.CL
In this paper we explore the task of modeling semi-structured object sequences; in particular, we focus our attention on the problem of developing a structure-aware input representation for such sequences. Examples of such data include user activity on websites, machine logs, and many others. This type of data is often...
http://arxiv.org/pdf/2301.01015v4
2023-05-23T02:33:22
2023-01-03T09:19:41
{ "title": [ -0.016525269, 0.005378723, -0.04232788, -0.080078125, -0.011650085, -0.01637268, -0.068725586, 0.019439697, 0.01335907, 0.059173584, -0.020202637, -0.010818481, 0.026428223, -0.032409668, -0.025817871, 0.024047852, 0.024673462, -0.00...
http://arxiv.org/abs/2301.00982v2
Analogical Inference Enhanced Knowledge Graph Embedding
Zhen Yao, Wen Zhang, Mingyang Chen, Yufeng Huang, Yi Yang, Huajun Chen
cs.AI, cs.CL
Knowledge graph embedding (KGE), which maps entities and relations in a knowledge graph into continuous vector spaces, has achieved great success in predicting missing links in knowledge graphs. However, knowledge graphs often contain incomplete triples that are difficult to inductively infer by KGEs. To address this c...
http://arxiv.org/pdf/2301.00982v2
2023-02-04T07:34:56
2023-01-03T07:24:05
{ "title": [ -0.009849548, -0.00058841705, -0.046325684, -0.05758667, -0.0031776428, -0.01348114, -0.025299072, -0.00064754486, 0.018569946, 0.012931824, -0.029342651, 0.010139465, 0.0135650635, -0.014282227, -0.015731812, 0.0066719055, -0.0005884170...
http://arxiv.org/abs/2301.00972v1
EZInterviewer: To Improve Job Interview Performance with Mock Interview Generator
Mingzhe Li, Xiuying Chen, Weiheng Liao, Yang Song, Tao Zhang, Dongyan Zhao, Rui Yan
cs.CL
Interview has been regarded as one of the most crucial step for recruitment. To fully prepare for the interview with the recruiters, job seekers usually practice with mock interviews between each other. However, such a mock interview with peers is generally far away from the real interview experience: the mock intervie...
http://arxiv.org/pdf/2301.00972v1
2023-01-03T07:00:30
2023-01-03T07:00:30
{ "title": [ 0.01033783, 0.011543274, -0.068603516, -0.017822266, 0.004096985, -0.03591919, 0.0039711, 0.005844116, -0.01550293, 0.018539429, 0.011688232, -0.032073975, 0.035339355, -0.07098389, 0.012077332, -0.027542114, 0.0090789795, 0.01834106...
http://arxiv.org/abs/2301.00818v1
ClusTop: An unsupervised and integrated text clustering and topic extraction framework
Zhongtao Chen, Chenghu Mi, Siwei Duo, Jingfei He, Yatong Zhou
cs.CL
Text clustering and topic extraction are two important tasks in text mining. Usually, these two tasks are performed separately. For topic extraction to facilitate clustering, we can first project texts into a topic space and then perform a clustering algorithm to obtain clusters. To promote topic extraction by clusteri...
http://arxiv.org/pdf/2301.00818v1
2023-01-03T03:26:26
2023-01-03T03:26:26
{ "title": [ -0.017562866, -0.0046463013, -0.049621582, -0.046722412, 0.018341064, 0.0056037903, -0.020111084, -0.0309906, 0.013748169, 0.025253296, -0.011108398, 0.041381836, -0.015136719, -0.006427765, -0.03668213, 0.017913818, -0.014640808, 0....
http://arxiv.org/abs/2301.00891v1
Understanding Political Polarisation using Language Models: A dataset and method
Samiran Gode, Supreeth Bare, Bhiksha Raj, Hyungon Yoo
cs.CL
Our paper aims to analyze political polarization in US political system using Language Models, and thereby help candidates make an informed decision. The availability of this information will help voters understand their candidates views on the economy, healthcare, education and other social issues. Our main contributi...
http://arxiv.org/pdf/2301.00891v1
2023-01-02T22:15:04
2023-01-02T22:15:04
{ "title": [ -0.00034284592, 0.0066070557, -0.02798462, -0.046783447, 0.013961792, -0.0070343018, -0.042785645, -0.026184082, -0.018875122, 0.035308838, -0.00680542, -0.012084961, 0.0049362183, -0.017196655, -0.059020996, -0.00040650368, -0.023101807...
http://arxiv.org/abs/2301.00876v2
MAUD: An Expert-Annotated Legal NLP Dataset for Merger Agreement Understanding
Steven H. Wang, Antoine Scardigli, Leonard Tang, Wei Chen, Dimitry Levkin, Anya Chen, Spencer Ball, Thomas Woodside, Oliver Zhang, Dan Hendrycks
cs.CL
Reading comprehension of legal text can be a particularly challenging task due to the length and complexity of legal clauses and a shortage of expert-annotated datasets. To address this challenge, we introduce the Merger Agreement Understanding Dataset (MAUD), an expert-annotated reading comprehension dataset based on ...
http://arxiv.org/pdf/2301.00876v2
2023-01-06T19:10:54
2023-01-02T21:08:27
{ "title": [ -0.048034668, -0.02557373, -0.061340332, -0.050445557, -0.015327454, -0.0015382767, -0.045196533, 0.030761719, 0.03149414, 0.010940552, -0.066833496, 0.013923645, 0.006790161, 0.004573822, -0.030593872, -0.011207581, -0.022979736, 0....
http://arxiv.org/abs/2301.00867v1
Follow the Timeline! Generating Abstractive and Extractive Timeline Summary in Chronological Order
Xiuying Chen, Mingzhe Li, Shen Gao, Zhangming Chan, Dongyan Zhao, Xin Gao, Xiangliang Zhang, Rui Yan
cs.CL
Nowadays, time-stamped web documents related to a general news query floods spread throughout the Internet, and timeline summarization targets concisely summarizing the evolution trajectory of events along the timeline. Unlike traditional document summarization, timeline summarization needs to model the time series inf...
http://arxiv.org/pdf/2301.00867v1
2023-01-02T20:29:40
2023-01-02T20:29:40
{ "title": [ -0.010604858, -0.011001587, -0.074523926, -0.09197998, -0.007255554, -0.015266418, -0.058410645, 0.035217285, -0.021408081, 0.036987305, -0.014228821, 0.032958984, -0.027938843, -0.029693604, -0.0021324158, -0.025436401, 0.012031555, ...
http://arxiv.org/abs/2301.00792v1
The Undesirable Dependence on Frequency of Gender Bias Metrics Based on Word Embeddings
Francisco Valentini, Germán Rosati, Diego Fernandez Slezak, Edgar Altszyler
cs.CL, cs.AI
Numerous works use word embedding-based metrics to quantify societal biases and stereotypes in texts. Recent studies have found that word embeddings can capture semantic similarity but may be affected by word frequency. In this work we study the effect of frequency when measuring female vs. male gender bias with word e...
http://arxiv.org/pdf/2301.00792v1
2023-01-02T18:27:10
2023-01-02T18:27:10
{ "title": [ -0.017791748, -0.006816864, -0.048797607, -0.020858765, 0.005592346, -0.031921387, -0.010398865, -0.011268616, 0.0046844482, 0.045654297, -0.047729492, -0.001244545, 0.030166626, 0.0023937225, -0.039978027, -0.0062561035, -0.02116394, ...
http://arxiv.org/abs/2301.00716v1
IRT2: Inductive Linking and Ranking in Knowledge Graphs of Varying Scale
Felix Hamann, Adrian Ulges, Maurice Falk
cs.LG, cs.AI, cs.CL
We address the challenge of building domain-specific knowledge models for industrial use cases, where labelled data and taxonomic information is initially scarce. Our focus is on inductive link prediction models as a basis for practical tools that support knowledge engineers with exploring text collections and discover...
http://arxiv.org/pdf/2301.00716v1
2023-01-02T15:19:21
2023-01-02T15:19:21
{ "title": [ -0.022460938, -0.005554199, -0.060668945, -0.05618286, 0.015975952, -0.00907135, -0.044158936, -0.017425537, -0.013473511, 0.006149292, -0.030548096, 0.01550293, -0.027572632, -0.0137786865, -0.017837524, -0.01914978, -0.0025901794, ...
http://arxiv.org/abs/2301.00709v1
Tsetlin Machine Embedding: Representing Words Using Logical Expressions
Bimal Bhattarai, Ole-Christoffer Granmo, Lei Jiao, Rohan Yadav, Jivitesh Sharma
cs.CL, cs.AI, cs.LG
Embedding words in vector space is a fundamental first step in state-of-the-art natural language processing (NLP). Typical NLP solutions employ pre-defined vector representations to improve generalization by co-locating similar words in vector space. For instance, Word2Vec is a self-supervised predictive model that cap...
http://arxiv.org/pdf/2301.00709v1
2023-01-02T15:02:45
2023-01-02T15:02:45
{ "title": [ -0.01574707, 0.0006055832, -0.042236328, -0.03652954, 0.001159668, -0.01071167, -0.037506104, -0.025985718, 0.0033493042, 0.032226562, -0.036376953, 0.0034809113, 0.02368164, -0.015640259, -0.026260376, 0.0017490387, -0.021575928, 0....
http://arxiv.org/abs/2301.00628v2
Using Active Learning Methods to Strategically Select Essays for Automated Scoring
Tahereh Firoozi, Hamid Mohammadi, Mark J. Gierl
cs.CL
Research on automated essay scoring has become increasing important because it serves as a method for evaluating students' written-responses at scale. Scalable methods for scoring written responses are needed as students migrate to online learning environments resulting in the need to evaluate large numbers of written-...
http://arxiv.org/pdf/2301.00628v2
2023-04-13T23:17:58
2023-01-02T12:46:10
{ "title": [ 0.008132935, 0.012641907, 0.006980896, 0.00014567375, 0.009353638, -0.021621704, 0.04852295, -0.0501709, -0.0050697327, 0.022079468, -0.014030457, 0.016860962, -0.049072266, -0.018234253, -0.037963867, -0.021392822, 0.0018491745, -0....
http://arxiv.org/abs/2301.00626v2
Design and analysis of tweet-based election models for the 2021 Mexican legislative election
Alejandro Vigna-Gómez, Javier Murillo, Manelik Ramirez, Alberto Borbolla, Ian Márquez, Prasun K. Ray
cs.SI, cs.CL, cs.CY
Modelling and forecasting real-life human behaviour using online social media is an active endeavour of interest in politics, government, academia, and industry. Since its creation in 2006, Twitter has been proposed as a potential laboratory that could be used to gauge and predict social behaviour. During the last deca...
http://arxiv.org/pdf/2301.00626v2
2023-06-21T08:01:38
2023-01-02T12:40:05
{ "title": [ -0.014442444, -0.005519867, -0.022964478, -0.04208374, -0.005355835, 0.00856781, -0.030075073, 0.004463196, -0.00021743774, 0.037902832, -0.022720337, 0.022659302, 0.010017395, -0.012802124, -0.030960083, 0.011581421, -0.013130188, 0...
http://arxiv.org/abs/2301.00604v1
Russia-Ukraine war: Modeling and Clustering the Sentiments Trends of Various Countries
Hamed Vahdat-Nejad, Mohammad Ghasem Akbari, Fatemeh Salmani, Faezeh Azizi, Hamid-Reza Nili-Sani
cs.CL
With Twitter's growth and popularity, a huge number of views are shared by users on various topics, making this platform a valuable information source on various political, social, and economic issues. This paper investigates English tweets on the Russia-Ukraine war to analyze trends reflecting users' opinions and sent...
http://arxiv.org/pdf/2301.00604v1
2023-01-02T11:32:47
2023-01-02T11:32:47
{ "title": [ 0.04083252, 0.004688263, -0.020736694, -0.035339355, 0.020065308, -0.026489258, 0.011802673, -0.008026123, -0.031677246, 0.04660034, -0.04888916, 0.02407837, 0.0065689087, 0.0006098747, -0.024917603, -0.018051147, 0.009185791, 0.0104...
http://arxiv.org/abs/2301.00591v3
Analysing Discrete Self Supervised Speech Representation for Spoken Language Modeling
Amitay Sicherman, Yossi Adi
cs.CL, cs.SD, eess.AS
This work profoundly analyzes discrete self-supervised speech representations (units) through the eyes of Generative Spoken Language Modeling (GSLM). Following the findings of such an analysis, we propose practical improvements to the discrete unit for the GSLM. First, we start comprehending these units by analyzing th...
http://arxiv.org/pdf/2301.00591v3
2023-03-01T09:59:54
2023-01-02T10:36:40
{ "title": [ -0.006881714, 0.008468628, -0.057495117, -0.07409668, 0.009498596, -0.015068054, -0.061187744, 0.025299072, -0.026885986, 0.037200928, -0.0059127808, -0.005302429, 0.00409317, -0.017196655, -0.05114746, -0.0025234222, -0.012001038, 0...
http://arxiv.org/abs/2301.01170v1
Transformer Based Geocoding
Yuval Solaz, Vitaly Shalumov
cs.CL, cs.AI
In this paper, we formulate the problem of predicting a geolocation from free text as a sequence-to-sequence problem. Using this formulation, we obtain a geocoding model by training a T5 encoder-decoder transformer model using free text as an input and geolocation as an output. The geocoding model was trained on geo-ta...
http://arxiv.org/pdf/2301.01170v1
2023-01-02T10:13:32
2023-01-02T10:13:32
{ "title": [ 0.031463623, 0.015289307, -0.0070114136, -0.01007843, -0.034698486, -0.017929077, -0.020401001, -0.035583496, -0.0012998581, 0.05014038, -0.042114258, -0.015342712, 0.05343628, -0.03036499, 0.008407593, 0.00806427, 0.027252197, 0.028...
http://arxiv.org/abs/2301.00565v1
Using meaning instead of words to track topics
Judicael Poumay, Ashwin Ittoo
cs.CL, cs.AI
The ability to monitor the evolution of topics over time is extremely valuable for businesses. Currently, all existing topic tracking methods use lexical information by matching word usage. However, no studies has ever experimented with the use of semantic information for tracking topics. Hence, we explore a novel sema...
http://arxiv.org/pdf/2301.00565v1
2023-01-02T08:55:55
2023-01-02T08:55:55
{ "title": [ -0.006668091, -0.0016803741, -0.025527954, -0.0059013367, -0.042297363, -0.014144897, -0.029937744, -0.029266357, -0.011001587, 0.031311035, -0.0049362183, -0.007408142, -0.005268097, -0.03567505, 0.036468506, -0.015258789, 0.020812988, ...
http://arxiv.org/abs/2301.00539v1
Statistical Machine Translation for Indic Languages
Sudhansu Bala Das, Divyajoti Panda, Tapas Kumar Mishra, Bidyut Kr. Patra
cs.CL
Machine Translation (MT) system generally aims at automatic representation of source language into target language retaining the originality of context using various Natural Language Processing (NLP) techniques. Among various NLP methods, Statistical Machine Translation(SMT). SMT uses probabilistic and statistical tech...
http://arxiv.org/pdf/2301.00539v1
2023-01-02T06:23:12
2023-01-02T06:23:12
{ "title": [ -0.046966553, -0.007583618, -0.0135650635, -0.0435791, -0.012535095, 0.016662598, -0.046417236, 0.010025024, -0.0107421875, 0.041992188, -0.016067505, 0.026535034, 0.004184723, -0.017440796, -0.035064697, 0.012512207, 0.006565094, 0....
http://arxiv.org/abs/2301.00503v3
A Concept Knowledge Graph for User Next Intent Prediction at Alipay
Yacheng He, Qianghuai Jia, Lin Yuan, Ruopeng Li, Yixin Ou, Ningyu Zhang
cs.CL, cs.AI, cs.DB, cs.IR, cs.LG
This paper illustrates the technologies of user next intent prediction with a concept knowledge graph. The system has been deployed on the Web at Alipay, serving more than 100 million daily active users. To explicitly characterize user intent, we propose AlipayKG, which is an offline concept knowledge graph in the Life...
http://arxiv.org/pdf/2301.00503v3
2023-03-14T11:01:26
2023-01-02T02:10:18
{ "title": [ -0.017227173, 0.0014657974, -0.06185913, -0.0413208, -0.0138168335, -0.029449463, -0.037628174, -0.008758545, 0.009613037, 0.04547119, -0.021957397, 0.013061523, 0.009140015, -0.007637024, -0.05114746, -0.004295349, -0.01965332, 0.01...
http://arxiv.org/abs/2301.00479v2
The Design Principle of Blockchain: An Initiative for the SoK of SoKs
Luyao Sunshine Zhang
cs.CR, cs.AI, cs.CL, stat.ML
Blockchain, also coined as decentralized AI, has the potential to empower AI to be more trustworthy by creating a decentralized trust of privacy, security, and audibility. However, systematic studies on the design principle of blockchain as a trust engine for an integrated society of cyber-physical-social-system (CPSS)...
http://arxiv.org/pdf/2301.00479v2
2023-01-14T21:18:01
2023-01-01T21:57:25
{ "title": [ -0.020812988, 0.0104599, -0.053955078, -0.014930725, 0.0037612915, -0.03427124, -0.010238647, -0.008491516, -0.040130615, 0.04043579, -0.0038661957, -0.0012216568, 0.022369385, -0.04244995, 0.004966736, -0.018096924, 0.013145447, 0.0...
http://arxiv.org/abs/2301.00429v1
Integrating Semantic Information into Sketchy Reading Module of Retro-Reader for Vietnamese Machine Reading Comprehension
Hang Thi-Thu Le, Viet-Duc Ho, Duc-Vu Nguyen, Ngan Luu-Thuy Nguyen
cs.CL
Machine Reading Comprehension has become one of the most advanced and popular research topics in the fields of Natural Language Processing in recent years. The classification of answerability questions is a relatively significant sub-task in machine reading comprehension; however, there haven't been many studies. Retro...
http://arxiv.org/pdf/2301.00429v1
2023-01-01T15:28:27
2023-01-01T15:28:27
{ "title": [ -0.04486084, -0.008796692, -0.02961731, -0.025024414, -0.03262329, 0.011268616, -0.050811768, -0.019195557, 0.007801056, 0.048187256, -0.029388428, 0.010391235, 0.0036830902, -0.025039673, -0.036590576, -0.0013999939, -0.006034851, 0...
http://arxiv.org/abs/2301.00422v1
Leveraging Semantic Representations Combined with Contextual Word Representations for Recognizing Textual Entailment in Vietnamese
Quoc-Loc Duong, Duc-Vu Nguyen, Ngan Luu-Thuy Nguyen
cs.CL
RTE is a significant problem and is a reasonably active research community. The proposed research works on the approach to this problem are pretty diverse with many different directions. For Vietnamese, the RTE problem is moderately new, but this problem plays a vital role in natural language understanding systems. Cur...
http://arxiv.org/pdf/2301.00422v1
2023-01-01T15:13:25
2023-01-01T15:13:25
{ "title": [ 0.016708374, -0.0008573532, -0.023132324, -0.07110596, -0.0044898987, 0.0031986237, -0.056549072, 0.005050659, 0.008712769, 0.03100586, -0.03765869, 0.020446777, 0.021331787, -0.0056533813, -0.032928467, 0.000992775, -0.023757935, 0....
http://arxiv.org/abs/2301.00418v1
Is word segmentation necessary for Vietnamese sentiment classification?
Duc-Vu Nguyen, Ngan Luu-Thuy Nguyen
cs.CL
To the best of our knowledge, this paper made the first attempt to answer whether word segmentation is necessary for Vietnamese sentiment classification. To do this, we presented five pre-trained monolingual S4- based language models for Vietnamese, including one model without word segmentation, and four models using R...
http://arxiv.org/pdf/2301.00418v1
2023-01-01T15:04:47
2023-01-01T15:04:47
{ "title": [ -0.0030536652, -0.007408142, -0.010559082, -0.012924194, -0.038848877, -0.004611969, -0.028274536, -0.013946533, -0.012756348, 0.033843994, -0.064331055, -0.008522034, -0.0037631989, -0.008918762, -0.042053223, -0.033691406, -0.009384155...
http://arxiv.org/abs/2301.00399v1
Semantic Operator Prediction and Applications
Farshad Noravesh
cs.CL
In the present paper, semantic parsing challenges are briefly introduced and QDMR formalism in semantic parsing is implemented using sequence to sequence model with attention but uses only part of speech(POS) as a representation of words of a sentence to make the training as simple and as fast as possible and also avoi...
http://arxiv.org/pdf/2301.00399v1
2023-01-01T13:20:57
2023-01-01T13:20:57
{ "title": [ -0.0046463013, 0.0030727386, -0.02696228, -0.07147217, 0.009994507, 0.003376007, -0.044891357, -0.02520752, 0.022277832, 0.03213501, -0.004348755, -0.020507812, 0.016204834, -0.025405884, -0.03540039, 0.026901245, 0.0016717911, 0.032...
http://arxiv.org/abs/2301.00397v1
Inflected Forms Are Redundant in Question Generation Models
Xingwu Sun, Hongyin Tang, chengzhong Xu
cs.CL
Neural models with an encoder-decoder framework provide a feasible solution to Question Generation (QG). However, after analyzing the model vocabulary we find that current models (both RNN-based and pre-training based) have more than 23\% inflected forms. As a result, the encoder will generate separate embeddings for t...
http://arxiv.org/pdf/2301.00397v1
2023-01-01T13:08:11
2023-01-01T13:08:11
{ "title": [ -0.003129959, -0.01550293, -0.026473999, -0.017974854, -0.04034424, -0.01979065, -0.013885498, 0.010284424, 0.017684937, 0.061431885, -0.00066137314, -0.037017822, 0.010871887, -0.039764404, -0.0287323, -0.0009813309, -0.0105896, 0.0...
http://arxiv.org/abs/2301.00395v1
CORGI-PM: A Chinese Corpus For Gender Bias Probing and Mitigation
Ge Zhang, Yizhi Li, Yaoyao Wu, Linyuan Zhang, Chenghua Lin, Jiayi Geng, Shi Wang, Jie Fu
cs.CL, cs.AI, cs.CY, cs.LG
As natural language processing (NLP) for gender bias becomes a significant interdisciplinary topic, the prevalent data-driven techniques such as large-scale language models suffer from data inadequacy and biased corpus, especially for languages with insufficient resources such as Chinese. To this end, we propose a Chin...
http://arxiv.org/pdf/2301.00395v1
2023-01-01T12:48:12
2023-01-01T12:48:12
{ "title": [ -0.0335083, -0.007095337, -0.04824829, -0.03744507, 0.004711151, -0.02078247, -0.022033691, -0.00042510033, 0.0128479, 0.04925537, -0.027832031, -0.0052871704, 0.053497314, -0.0070648193, -0.04550171, 0.0047836304, -0.036590576, 0.06...
http://arxiv.org/abs/2301.00374v1
Optimizing Readability Using Genetic Algorithms
Jorge Martinez-Gil
cs.CL, cs.AI
This research presents ORUGA, a method that tries to automatically optimize the readability of any text in English. The core idea behind the method is that certain factors affect the readability of a text, some of which are quantifiable (number of words, syllables, presence or absence of adverbs, and so on). The nature...
http://arxiv.org/pdf/2301.00374v1
2023-01-01T09:08:45
2023-01-01T09:08:45
{ "title": [ -0.0513916, -0.004295349, -0.017456055, -0.017669678, -0.015182495, 0.014984131, -0.03552246, -0.03125, 0.000813961, 0.05227661, -0.0061302185, -0.019546509, -0.00793457, -0.013198853, -0.06616211, -0.0026950836, -0.00894928, 0.01415...
http://arxiv.org/abs/2301.00355v2
Second Thoughts are Best: Learning to Re-Align With Human Values from Text Edits
Ruibo Liu, Chenyan Jia, Ge Zhang, Ziyu Zhuang, Tony X Liu, Soroush Vosoughi
cs.CL, cs.AI, cs.CY
We present Second Thought, a new learning paradigm that enables language models (LMs) to re-align with human values. By modeling the chain-of-edits between value-unaligned and value-aligned text, with LM fine-tuning and additional refinement through reinforcement learning, Second Thought not only achieves superior perf...
http://arxiv.org/pdf/2301.00355v2
2023-01-05T00:17:12
2023-01-01T05:56:21
{ "title": [ -0.01914978, -0.009864807, -0.011116028, -0.004245758, -0.018844604, -0.011383057, -0.011581421, -0.020889282, -0.015899658, 0.0357666, -0.014152527, 0.012680054, 0.008255005, -0.0046653748, -0.04748535, -0.0074768066, -0.010314941, ...
http://arxiv.org/abs/2301.01743v1
Chatbots as Problem Solvers: Playing Twenty Questions with Role Reversals
David Noever, Forrest McKee
cs.AI, cs.CL
New chat AI applications like ChatGPT offer an advanced understanding of question context and memory across multi-step tasks, such that experiments can test its deductive reasoning. This paper proposes a multi-role and multi-step challenge, where ChatGPT plays the classic twenty-questions game but innovatively switches...
http://arxiv.org/pdf/2301.01743v1
2023-01-01T03:04:04
2023-01-01T03:04:04
{ "title": [ 0.0026626587, -0.017105103, -0.014640808, -0.0181427, -0.020767212, -0.010520935, 0.006881714, -0.024108887, -0.015235901, 0.024719238, -0.012519836, -0.018096924, -0.012527466, -0.02355957, -0.033599854, -0.024002075, -0.0057868958, ...
http://arxiv.org/abs/2301.00321v1
Floods Relevancy and Identification of Location from Twitter Posts using NLP Techniques
Muhammad Suleman, Muhammad Asif, Tayyab Zamir, Ayaz Mehmood, Jebran Khan, Nasir Ahmad, Kashif Ahmad
cs.CL
This paper presents our solutions for the MediaEval 2022 task on DisasterMM. The task is composed of two subtasks, namely (i) Relevance Classification of Twitter Posts (RCTP), and (ii) Location Extraction from Twitter Texts (LETT). The RCTP subtask aims at differentiating flood-related and non-relevant social posts whi...
http://arxiv.org/pdf/2301.00321v1
2023-01-01T01:36:32
2023-01-01T01:36:32
{ "title": [ 0.023651123, 0.011375427, -0.024307251, -0.039886475, 0.022720337, -0.01499939, 0.0017719269, -0.030212402, 0.0002593994, 0.058013916, -0.041412354, 0.011329651, 0.035583496, 0.00089120865, -0.047851562, -0.00032234192, -0.023590088, ...
http://arxiv.org/abs/2301.00320v1
Relevance Classification of Flood-related Twitter Posts via Multiple Transformers
Wisal Mukhtiar, Waliiya Rizwan, Aneela Habib, Yasir Saleem Afridi, Laiq Hasan, Kashif Ahmad
cs.CL
In recent years, social media has been widely explored as a potential source of communication and information in disasters and emergency situations. Several interesting works and case studies of disaster analytics exploring different aspects of natural disasters have been already conducted. Along with the great potenti...
http://arxiv.org/pdf/2301.00320v1
2023-01-01T01:34:15
2023-01-01T01:34:15
{ "title": [ 0.021026611, 0.016479492, -0.02281189, -0.068481445, 0.010910034, 0.0016622543, -0.033569336, -0.0029296875, -0.011413574, 0.05441284, -0.043304443, 0.020874023, 0.01651001, 0.0026950836, -0.064941406, -0.01108551, -0.01525116, 0.034...
http://arxiv.org/abs/2301.00304v1
Sample-Efficient Unsupervised Domain Adaptation of Speech Recognition Systems A case study for Modern Greek
Georgios Paraskevopoulos, Theodoros Kouzelis, Georgios Rouvalis, Athanasios Katsamanis, Vassilis Katsouros, Alexandros Potamianos
cs.CL, cs.SD, eess.AS
Modern speech recognition systems exhibits rapid performance degradation under domain shift. This issue is especially prevalent in data-scarce settings, such as low-resource languages, where diversity of training data is limited. In this work we propose M2DS2, a simple and sample-efficient finetuning strategy for large...
http://arxiv.org/pdf/2301.00304v1
2022-12-31T22:57:30
2022-12-31T22:57:30
{ "title": [ 0.0011014938, 0.008773804, -0.03338623, -0.055023193, 0.016662598, -0.013946533, -0.02482605, 0.010978699, -0.0031795502, 0.03753662, -0.039794922, 0.0022277832, 0.015213013, -0.011299133, -0.041107178, -0.0046653748, 0.003232956, 0....
http://arxiv.org/abs/2301.00303v1
Rethinking with Retrieval: Faithful Large Language Model Inference
Hangfeng He, Hongming Zhang, Dan Roth
cs.CL, cs.AI
Despite the success of large language models (LLMs) in various natural language processing (NLP) tasks, the stored knowledge in these models may inevitably be incomplete, out-of-date, or incorrect. This motivates the need to utilize external knowledge to assist LLMs. Unfortunately, current methods for incorporating ext...
http://arxiv.org/pdf/2301.00303v1
2022-12-31T22:35:34
2022-12-31T22:35:34
{ "title": [ -0.017318726, -0.014717102, -0.034820557, -0.076660156, 0.005882263, 0.0063667297, -0.06262207, -0.003929138, 0.016677856, 0.026672363, -0.019699097, -0.0042381287, 0.008605957, -0.017562866, -0.04486084, 0.015144348, -0.004585266, 0...
http://arxiv.org/abs/2301.00234v3
A Survey on In-context Learning
Qingxiu Dong, Lei Li, Damai Dai, Ce Zheng, Zhiyong Wu, Baobao Chang, Xu Sun, Jingjing Xu, Lei Li, Zhifang Sui
cs.CL, cs.AI
With the increasing ability of large language models (LLMs), in-context learning (ICL) has become a new paradigm for natural language processing (NLP), where LLMs make predictions only based on contexts augmented with a few examples. It has been a new trend to explore ICL to evaluate and extrapolate the ability of LLMs...
http://arxiv.org/pdf/2301.00234v3
2023-06-01T12:23:40
2022-12-31T15:57:09
{ "title": [ 0.01675415, 0.00040841103, -0.03741455, -0.04699707, 0.029953003, -0.03289795, 0.009437561, -0.012809753, 0.003894806, 0.036865234, -0.042022705, 0.009941101, -0.023635864, -0.0041503906, -0.036987305, -0.009346008, -0.002910614, 0.0...
http://arxiv.org/abs/2301.00200v2
Logic Mill -- A Knowledge Navigation System
Sebastian Erhardt, Mainak Ghosh, Erik Buunk, Michael E. Rose, Dietmar Harhoff
cs.CL
Logic Mill is a scalable and openly accessible software system that identifies semantically similar documents within either one domain-specific corpus or multi-domain corpora. It uses advanced Natural Language Processing (NLP) techniques to generate numerical representations of documents. Currently it leverages a large...
http://arxiv.org/pdf/2301.00200v2
2023-10-20T10:00:03
2022-12-31T13:46:50
{ "title": [ -0.022232056, -0.020248413, -0.0569458, -0.048614502, -0.024673462, -0.023895264, -0.033477783, 0.006008148, -0.006088257, 0.032592773, -0.037628174, -0.0103302, 0.0010204315, -0.03933716, -0.030197144, -0.016799927, 0.0068130493, 0....
http://arxiv.org/abs/2301.00152v1
Towards Proactively Forecasting Sentence-Specific Information Popularity within Online News Documents
Sayar Ghosh Roy, Anshul Padhi, Risubh Jain, Manish Gupta, Vasudeva Varma
cs.CL, cs.AI, cs.LG
Multiple studies have focused on predicting the prospective popularity of an online document as a whole, without paying attention to the contributions of its individual parts. We introduce the task of proactively forecasting popularities of sentences within online news documents solely utilizing their natural language ...
http://arxiv.org/pdf/2301.00152v1
2022-12-31T08:40:08
2022-12-31T08:40:08
{ "title": [ 0.0021839142, -0.002412796, -0.0184021, -0.051574707, -0.001455307, -0.0138549805, -0.009742737, -0.0025806427, 0.019561768, 0.047180176, -0.032928467, 0.009300232, 0.021850586, 0.002811432, -0.06011963, 0.0072288513, -0.021575928, 0...
http://arxiv.org/abs/2301.00068v2
On the Inconsistencies of Conditionals Learned by Masked Language Models
Tom Young, Yang You
cs.CL, cs.AI
Learning to predict masked tokens in a sequence has been shown to be a powerful pretraining objective for large language models. After training, such masked language models can provide distributions of tokens conditioned on bidirectional context. In this paper, we show that contrary to popular assumptions, such bidir...
http://arxiv.org/pdf/2301.00068v2
2023-10-08T07:57:03
2022-12-30T22:53:25
{ "title": [ -0.05847168, 0.008117676, -0.061035156, -0.043151855, -0.0016536713, -0.0008363724, -0.06222534, 0.007675171, -0.0039634705, 0.011138916, -0.044647217, -0.0070266724, 0.0035209656, -0.020202637, -0.030929565, -0.018829346, -0.028045654, ...
http://arxiv.org/abs/2301.00066v1
Memory Augmented Lookup Dictionary based Language Modeling for Automatic Speech Recognition
Yukun Feng, Ming Tu, Rui Xia, Chuanzeng Huang, Yuxuan Wang
cs.CL, eess.AS
Recent studies have shown that using an external Language Model (LM) benefits the end-to-end Automatic Speech Recognition (ASR). However, predicting tokens that appear less frequently in the training set is still quite challenging. The long-tail prediction problems have been widely studied in many applications, but onl...
http://arxiv.org/pdf/2301.00066v1
2022-12-30T22:26:57
2022-12-30T22:26:57
{ "title": [ -0.032928467, 0.00004631281, -0.044921875, -0.066467285, -0.005908966, 0.01423645, -0.06536865, 0.02633667, 0.0017728806, 0.04711914, -0.027252197, -0.011650085, 0.017562866, -0.012023926, -0.06274414, 0.001461029, -0.009239197, 0.03...
http://arxiv.org/abs/2212.14882v1
ChatGPT Makes Medicine Easy to Swallow: An Exploratory Case Study on Simplified Radiology Reports
Katharina Jeblick, Balthasar Schachtner, Jakob Dexl, Andreas Mittermeier, Anna Theresa Stüber, Johanna Topalis, Tobias Weber, Philipp Wesp, Bastian Sabel, Jens Ricke, Michael Ingrisch
cs.CL, cs.LG
The release of ChatGPT, a language model capable of generating text that appears human-like and authentic, has gained significant attention beyond the research community. We expect that the convincing performance of ChatGPT incentivizes users to apply it to a variety of downstream tasks, including prompting the model t...
http://arxiv.org/pdf/2212.14882v1
2022-12-30T18:55:16
2022-12-30T18:55:16
{ "title": [ 0.0019330978, -0.012893677, -0.030151367, -0.037750244, -0.018539429, -0.006629944, -0.029525757, -0.0023536682, -0.014129639, 0.02229309, -0.018859863, 0.010040283, -0.024627686, -0.035308838, -0.015289307, -0.008140564, -0.016708374, ...
http://arxiv.org/abs/2301.00688v1
Active Learning for Neural Machine Translation
Neeraj Vashistha, Kriti Singh, Ramakant Shakya
cs.CL, cs.AI
The machine translation mechanism translates texts automatically between different natural languages, and Neural Machine Translation (NMT) has gained attention for its rational context analysis and fluent translation accuracy. However, processing low-resource languages that lack relevant training attributes like superv...
http://arxiv.org/pdf/2301.00688v1
2022-12-30T17:04:01
2022-12-30T17:04:01
{ "title": [ -0.03225708, 0.0014238358, 0.005947113, -0.04864502, 0.0057525635, 0.016342163, -0.022659302, -0.0076446533, 0.005279541, 0.06011963, 0.011108398, 0.00028181076, 0.00356102, -0.026977539, -0.04336548, 0.03366089, 0.0068740845, 0.0195...
http://arxiv.org/abs/2212.14815v3
Black-box language model explanation by context length probing
Ondřej Cífka, Antoine Liutkus
cs.CL, cs.LG
The increasingly widespread adoption of large language models has highlighted the need for improving their explainability. We present context length probing, a novel explanation technique for causal language models, based on tracking the predictions of a model as a function of the length of available context, and allow...
http://arxiv.org/pdf/2212.14815v3
2023-05-26T20:32:25
2022-12-30T16:24:10
{ "title": [ -0.03829956, -0.02418518, -0.03378296, -0.06439209, -0.020004272, 0.0070114136, -0.050872803, 0.006893158, -0.019302368, 0.022399902, -0.014953613, 0.0069732666, -0.0023899078, -0.019088745, -0.023757935, -0.00001758337, -0.01374054, ...
http://arxiv.org/abs/2301.06902v1
TA-DA: Topic-Aware Domain Adaptation for Scientific Keyphrase Identification and Classification (Student Abstract)
Răzvan-Alexandru Smădu, George-Eduard Zaharia, Andrei-Marius Avram, Dumitru-Clementin Cercel, Mihai Dascalu, Florin Pop
cs.CL
Keyphrase identification and classification is a Natural Language Processing and Information Retrieval task that involves extracting relevant groups of words from a given text related to the main topic. In this work, we focus on extracting keyphrases from scientific documents. We introduce TA-DA, a Topic-Aware Domain A...
http://arxiv.org/pdf/2301.06902v1
2022-12-30T15:07:27
2022-12-30T15:07:27
{ "title": [ -0.04586792, 0.0032367706, -0.046295166, -0.0034122467, -0.0068473816, -0.011711121, -0.0018262863, -0.010414124, 0.010383606, 0.062347412, 0.039276123, 0.009468079, 0.010246277, -0.026046753, -0.0569458, 0.0309906, -0.0317688, 0.036...
http://arxiv.org/abs/2212.14657v1
Linear programming word problems formulation using EnsembleCRF NER labeler and T5 text generator with data augmentations
JiangLong He, Mamatha N, Shiv Vignesh, Deepak Kumar, Akshay Uppal
cs.CL, cs.AI
We propose an ensemble approach to predict the labels in linear programming word problems. The entity identification and the meaning representation are two types of tasks to be solved in the NL4Opt competition. We propose the ensembleCRF method to identify the named entities for the first task. We found that single mod...
http://arxiv.org/pdf/2212.14657v1
2022-12-30T12:39:26
2022-12-30T12:39:26
{ "title": [ -0.028182983, -0.024017334, -0.038024902, -0.0028457642, -0.05041504, -0.0184021, -0.044891357, 0.0054244995, -0.04660034, -0.0047454834, -0.03564453, 0.034332275, -0.026351929, -0.03488159, 0.036376953, -0.029769897, 0.0029830933, -...
http://arxiv.org/abs/2212.14648v1
Distant Reading of the German Coalition Deal: Recognizing Policy Positions with BERT-based Text Classification
Michael Zylla, Thomas Haider
cs.CL
Automated text analysis has become a widely used tool in political science. In this research, we use a BERT model trained on German party manifestos to identify the individual parties' contribution to the coalition agreement of 2021.
http://arxiv.org/pdf/2212.14648v1
2022-12-30T12:20:39
2022-12-30T12:20:39
{ "title": [ -0.0057373047, -0.007698059, -0.0079193115, -0.020507812, -0.026870728, -0.00919342, -0.0317688, -0.012107849, 0.0020008087, 0.030471802, -0.030807495, 0.015266418, 0.0028877258, -0.0184021, -0.042022705, -0.03515625, -0.019836426, 0...
http://arxiv.org/abs/2212.14578v1
MAUVE Scores for Generative Models: Theory and Practice
Krishna Pillutla, Lang Liu, John Thickstun, Sean Welleck, Swabha Swayamdipta, Rowan Zellers, Sewoong Oh, Yejin Choi, Zaid Harchaoui
cs.LG, cs.AI, cs.CL
Generative AI has matured to a point where large-scale models can generate text that seems indistinguishable from human-written text and remarkably photorealistic images. Automatically measuring how close the distribution of generated data is to the target real data distribution is a key step in diagnosing existing mod...
http://arxiv.org/pdf/2212.14578v1
2022-12-30T07:37:40
2022-12-30T07:37:40
{ "title": [ -0.035186768, -0.009384155, -0.040802002, -0.045898438, -0.009613037, 0.004535675, -0.052459717, -0.014266968, 0.018554688, 0.0031490326, 0.011726379, 0.0045547485, -0.005340576, -0.045440674, -0.023635864, 0.01940918, -0.006385803, ...
http://arxiv.org/abs/2212.14548v3
How would Stance Detection Techniques Evolve after the Launch of ChatGPT?
Bowen Zhang, Daijun Ding, Liwen Jing
cs.CL
Stance detection refers to the task of extracting the standpoint (Favor, Against or Neither) towards a target in given texts. Such research gains increasing attention with the proliferation of social media contents. The conventional framework of handling stance detection is converting it into text classification tasks....
http://arxiv.org/pdf/2212.14548v3
2023-04-10T11:43:12
2022-12-30T05:03:15
{ "title": [ -0.023117065, -0.023086548, -0.010940552, 0.021133423, -0.038269043, 0.0072135925, -0.0112838745, -0.010467529, 0.023529053, 0.037506104, -0.035949707, -0.0011997223, -0.004009247, -0.0037727356, -0.027709961, -0.014892578, 0.029296875, ...
http://arxiv.org/abs/2212.14546v1
HiTeA: Hierarchical Temporal-Aware Video-Language Pre-training
Qinghao Ye, Guohai Xu, Ming Yan, Haiyang Xu, Qi Qian, Ji Zhang, Fei Huang
cs.CV, cs.CL, cs.MM
Video-language pre-training has advanced the performance of various downstream video-language tasks. However, most previous methods directly inherit or adapt typical image-language pre-training paradigms to video-language pre-training, thus not fully exploiting the unique characteristic of video, i.e., temporal. In thi...
http://arxiv.org/pdf/2212.14546v1
2022-12-30T04:27:01
2022-12-30T04:27:01
{ "title": [ -0.008682251, -0.0067977905, -0.06311035, -0.10748291, 0.009994507, -0.0056114197, -0.047973633, 0.025177002, -0.0143585205, 0.035003662, -0.0107421875, 0.014709473, 0.015640259, -0.015594482, -0.02470398, -0.004058838, -0.006549835, ...