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/2301.06009v1
Rationalizing Predictions by Adversarial Information Calibration
Lei Sha, Oana-Maria Camburu, Thomas Lukasiewicz
cs.CL, cs.AI
Explaining the predictions of AI models is paramount in safety-critical applications, such as in legal or medical domains. One form of explanation for a prediction is an extractive rationale, i.e., a subset of features of an instance that lead the model to give its prediction on that instance. For example, the subphras...
http://arxiv.org/pdf/2301.06009v1
2023-01-15T03:13:09
2023-01-15T03:13:09
{ "title": [ -0.03643799, 0.0028133392, -0.020706177, -0.043182373, -0.0064811707, 0.015533447, -0.036712646, -0.029968262, -0.024993896, 0.0335083, -0.00031781197, 0.025299072, -0.038024902, -0.008926392, -0.045959473, -0.011062622, 0.021347046, ...
http://arxiv.org/abs/2301.05948v3
tasksource: A Dataset Harmonization Framework for Streamlined NLP Multi-Task Learning and Evaluation
Damien Sileo
cs.CL, cs.AI, I.2.7
The HuggingFace Datasets Hub hosts thousands of datasets, offering exciting opportunities for language model training and evaluation. However, datasets for a specific task type often have different schemas, making harmonization challenging. Multi-task training or evaluation necessitates manual work to fit data into tas...
http://arxiv.org/pdf/2301.05948v3
2023-05-16T08:19:44
2023-01-14T16:38:04
{ "title": [ -0.014877319, -0.005168915, -0.06890869, -0.08111572, 0.015136719, -0.013046265, -0.051605225, 0.013145447, 0.0010080338, 0.032806396, -0.022872925, 0.024032593, -0.019851685, 0.0077209473, -0.058044434, -0.004878998, -0.028427124, 0...
http://arxiv.org/abs/2301.05880v3
TikTalk: A Video-Based Dialogue Dataset for Multi-Modal Chitchat in Real World
Hongpeng Lin, Ludan Ruan, Wenke Xia, Peiyu Liu, Jingyuan Wen, Yixin Xu, Di Hu, Ruihua Song, Wayne Xin Zhao, Qin Jin, Zhiwu Lu
cs.CL, cs.AI
To facilitate the research on intelligent and human-like chatbots with multi-modal context, we introduce a new video-based multi-modal dialogue dataset, called TikTalk. We collect 38K videos from a popular video-sharing platform, along with 367K conversations posted by users beneath them. Users engage in spontaneous co...
http://arxiv.org/pdf/2301.05880v3
2023-09-08T10:03:16
2023-01-14T10:18:22
{ "title": [ 0.005897522, -0.0022277832, -0.03933716, -0.057861328, 0.016586304, -0.01689148, -0.033721924, -0.00014841557, 0.00995636, 0.039123535, -0.018432617, -0.010116577, 0.01600647, -0.012145996, -0.050598145, -0.0043525696, -0.018859863, ...
http://arxiv.org/abs/2301.05863v1
Detecting Stance of Authorities towards Rumors in Arabic Tweets: A Preliminary Study
Fatima Haouari, Tamer Elsayed
cs.CL, cs.AI, cs.IR, cs.SI
A myriad of studies addressed the problem of rumor verification in Twitter by either utilizing evidence from the propagation networks or external evidence from the Web. However, none of these studies exploited evidence from trusted authorities. In this paper, we define the task of detecting the stance of authorities to...
http://arxiv.org/pdf/2301.05863v1
2023-01-14T09:28:43
2023-01-14T09:28:43
{ "title": [ -0.0014371872, 0.0034561157, -0.017578125, -0.029403687, 0.00075531006, -0.011856079, -0.005809784, -0.014846802, -0.018859863, 0.031036377, -0.04421997, 0.025817871, -0.016296387, -0.007713318, -0.04159546, -0.02067566, -0.024414062, ...
http://arxiv.org/abs/2301.05843v2
Leveraging Large Language Models to Power Chatbots for Collecting User Self-Reported Data
Jing Wei, Sungdong Kim, Hyunhoon Jung, Young-Ho Kim
cs.HC, cs.AI, cs.CL, H.5.2; I.2.7
Large language models (LLMs) provide a new way to build chatbots by accepting natural language prompts. Yet, it is unclear how to design prompts to power chatbots to carry on naturalistic conversations while pursuing a given goal, such as collecting self-report data from users. We explore what design factors of prompts...
http://arxiv.org/pdf/2301.05843v2
2023-09-22T22:33:24
2023-01-14T07:29:36
{ "title": [ 0.0008111, -0.012680054, -0.035827637, -0.04940796, -0.006252289, -0.008460999, -0.023330688, 0.008552551, -0.012786865, 0.03955078, -0.016494751, 0.006931305, -0.0032215118, -0.010414124, -0.047821045, 0.0035057068, -0.0027046204, 0...
http://arxiv.org/abs/2301.08145v1
Music Playlist Title Generation Using Artist Information
Haven Kim, SeungHeon Doh, Junwon Lee, Juhan Nam
cs.IR, cs.CL, cs.LG, cs.SD, eess.AS
Automatically generating or captioning music playlist titles given a set of tracks is of significant interest in music streaming services as customized playlists are widely used in personalized music recommendation, and well-composed text titles attract users and help their music discovery. We present an encoder-decode...
http://arxiv.org/pdf/2301.08145v1
2023-01-14T00:19:39
2023-01-14T00:19:39
{ "title": [ -0.03050232, -0.004886627, -0.062438965, -0.016220093, -0.028869629, -0.03314209, -0.014328003, 0.0016565323, -0.021850586, 0.0491333, -0.021331787, 0.038848877, 0.017669678, -0.022964478, -0.006313324, 0.015823364, -0.0031719208, 0....
http://arxiv.org/abs/2301.05746v1
Infusing Commonsense World Models with Graph Knowledge
Alexander Gurung, Mojtaba Komeili, Arthur Szlam, Jason Weston, Jack Urbanek
cs.CL, cs.AI
While language models have become more capable of producing compelling language, we find there are still gaps in maintaining consistency, especially when describing events in a dynamically changing world. We study the setting of generating narratives in an open world text adventure game, where a graph representation of...
http://arxiv.org/pdf/2301.05746v1
2023-01-13T19:58:27
2023-01-13T19:58:27
{ "title": [ -0.019119263, -0.040130615, -0.048065186, -0.06512451, -0.027816772, -0.0020828247, -0.04232788, 0.016662598, 0.016052246, 0.009353638, -0.02885437, 0.0067634583, -0.022094727, -0.021133423, -0.021057129, -0.004432678, 0.008888245, 0...
http://arxiv.org/abs/2301.05663v1
Natural Language Processing of Aviation Occurrence Reports for Safety Management
Patrick Jonk, Vincent de Vries, Rombout Wever, Georgios Sidiropoulos, Evangelos Kanoulas
cs.CL, cs.LG
Occurrence reporting is a commonly used method in safety management systems to obtain insight in the prevalence of hazards and accident scenarios. In support of safety data analysis, reports are often categorized according to a taxonomy. However, the processing of the reports can require significant effort from safety ...
http://arxiv.org/pdf/2301.05663v1
2023-01-13T17:00:09
2023-01-13T17:00:09
{ "title": [ 0.014205933, 0.0051498413, -0.04019165, -0.09509277, 0.019989014, -0.008117676, -0.039001465, 0.0104599, -0.039916992, 0.029525757, -0.03604126, 0.023925781, -0.009971619, -0.020355225, 0.008132935, -0.021743774, -0.015296936, 0.0302...
http://arxiv.org/abs/2301.05659v1
From stage to page: language independent bootstrap measures of distinctiveness in fictional speech
Artjoms Šeļa, Ben Nagy, Joanna Byszuk, Laura Hernández-Lorenzo, Botond Szemes, Maciej Eder
cs.CL
Stylometry is mostly applied to authorial style. Recently, researchers have begun investigating the style of characters, finding that the variation remains within authorial bounds. We address the stylistic distinctiveness of characters in drama. Our primary contribution is methodological; we introduce and evaluate two ...
http://arxiv.org/pdf/2301.05659v1
2023-01-13T16:58:43
2023-01-13T16:58:43
{ "title": [ -0.021347046, 0.0071258545, -0.057556152, -0.044525146, 0.010482788, -0.027633667, -0.013916016, 0.02998352, -0.021438599, 0.025222778, -0.00032806396, 0.02230835, 0.019042969, -0.041748047, -0.015266418, 0.011436462, -0.027023315, 0...
http://arxiv.org/abs/2301.05571v2
The 2022 n2c2/UW Shared Task on Extracting Social Determinants of Health
Kevin Lybarger, Meliha Yetisgen, Özlem Uzuner
cs.CL, I.2.7
Objective: The n2c2/UW SDOH Challenge explores the extraction of social determinant of health (SDOH) information from clinical notes. The objectives include the advancement of natural language processing (NLP) information extraction techniques for SDOH and clinical information more broadly. This paper presents the shar...
http://arxiv.org/pdf/2301.05571v2
2023-02-13T20:08:09
2023-01-13T14:20:23
{ "title": [ 0.009567261, 0.009742737, -0.03918457, -0.081604004, 0.022735596, -0.022766113, -0.050567627, -0.025894165, -0.03466797, 0.008728027, 0.02130127, 0.02810669, -0.028427124, -0.04663086, 0.029037476, 0.019348145, 0.0026836395, 0.015411...
http://arxiv.org/abs/2301.05562v1
Multilingual Alzheimer's Dementia Recognition through Spontaneous Speech: a Signal Processing Grand Challenge
Saturnino Luz, Fasih Haider, Davida Fromm, Ioulietta Lazarou, Ioannis Kompatsiaris, Brian MacWhinney
eess.AS, cs.AI, cs.CL, cs.LG, 68T10 (Primary) 92C55 (Secondary), J.3; I.2.6; I.5.1
This Signal Processing Grand Challenge (SPGC) targets a difficult automatic prediction problem of societal and medical relevance, namely, the detection of Alzheimer's Dementia (AD). Participants were invited to employ signal processing and machine learning methods to create predictive models based on spontaneous speech...
http://arxiv.org/pdf/2301.05562v1
2023-01-13T14:09:13
2023-01-13T14:09:13
{ "title": [ 0.008018494, 0.0042381287, -0.03161621, -0.060516357, 0.0054359436, -0.0025348663, -0.050079346, -0.0010242462, -0.015777588, 0.040740967, -0.024139404, -0.006198883, 0.0009431839, -0.019973755, -0.03842163, -0.024124146, -0.0071754456, ...
http://arxiv.org/abs/2301.05543v1
Classification of Cross-cultural News Events
Abdul Sittar, Dunja Mladenic
cs.CL, cs.AI
We present a methodology to support the analysis of culture from text such as news events and demonstrate its usefulness on categorizing news events from different categories (society, business, health, recreation, science, shopping, sports, arts, computers, games and home) across different geographical locations (diff...
http://arxiv.org/pdf/2301.05543v1
2023-01-13T13:41:18
2023-01-13T13:41:18
{ "title": [ 0.00969696, 0.018096924, -0.014717102, -0.03253174, 0.00079631805, -0.024017334, -0.024307251, -0.021621704, -0.03488159, 0.041748047, -0.0010128021, 0.0137786865, -0.003370285, -0.033996582, -0.0013561249, 0.008003235, 0.006038666, ...
http://arxiv.org/abs/2301.05494v1
Multilingual Detection of Check-Worthy Claims using World Languages and Adapter Fusion
Ipek Baris Schlicht, Lucie Flek, Paolo Rosso
cs.CL, cs.IR
Check-worthiness detection is the task of identifying claims, worthy to be investigated by fact-checkers. Resource scarcity for non-world languages and model learning costs remain major challenges for the creation of models supporting multilingual check-worthiness detection. This paper proposes cross-training adapters ...
http://arxiv.org/pdf/2301.05494v1
2023-01-13T11:50:08
2023-01-13T11:50:08
{ "title": [ 0.012145996, 0.008087158, -0.031463623, -0.07299805, 0.018203735, -0.011108398, -0.04827881, -0.019760132, -0.0395813, 0.045806885, -0.002407074, 0.00086069107, 0.028656006, -0.02810669, -0.004058838, -0.006690979, -0.022277832, 0.06...
http://arxiv.org/abs/2301.05487v1
Improving Generalization of Adapter-Based Cross-lingual Transfer with Scheduled Unfreezing
Chen Cecilia Liu, Jonas Pfeiffer, Ivan Vulić, Iryna Gurevych
cs.CL
Standard fine-tuning of language models typically performs well on in-distribution data, but suffers with generalization to distribution shifts. In this work, we aim to improve generalization of adapter-based cross-lingual task transfer where such cross-language distribution shifts are imminent. We investigate schedule...
http://arxiv.org/pdf/2301.05487v1
2023-01-13T11:26:53
2023-01-13T11:26:53
{ "title": [ -0.006340027, 0.03692627, -0.022521973, -0.04348755, 0.024749756, 0.0107803345, -0.040008545, -0.037475586, 0.002576828, 0.064331055, 0.003850937, 0.019165039, 0.047332764, -0.025985718, -0.018417358, 0.04083252, 0.0042266846, 0.0409...
http://arxiv.org/abs/2301.05478v1
Structuring ontologies in a context of collaborative system modelling
Romy Lynn Chaib, Rallou Thomopoulos, Catherine Macombe
cs.AI, cs.CL
Prospective studies require discussing and collaborating with the stakeholders to create scenarios of the possible evolution of the studied value-chain. However, stakeholders don't always use the same words when referring to one idea. Constructing an ontology and homogenizing vocabularies is thus crucial to identify ke...
http://arxiv.org/pdf/2301.05478v1
2023-01-13T11:13:39
2023-01-13T11:13:39
{ "title": [ -0.03955078, -0.0046081543, -0.06274414, -0.08068848, 0.016693115, 0.003440857, -0.060150146, 0.0017089844, -0.01373291, 0.024383545, -0.009521484, -0.0063438416, -0.03729248, -0.02178955, -0.043273926, -0.014770508, -0.015007019, 0....
http://arxiv.org/abs/2301.05453v2
It's Just a Matter of Time: Detecting Depression with Time-Enriched Multimodal Transformers
Ana-Maria Bucur, Adrian Cosma, Paolo Rosso, Liviu P. Dinu
cs.CL
Depression detection from user-generated content on the internet has been a long-lasting topic of interest in the research community, providing valuable screening tools for psychologists. The ubiquitous use of social media platforms lays out the perfect avenue for exploring mental health manifestations in posts and int...
http://arxiv.org/pdf/2301.05453v2
2023-02-06T14:42:24
2023-01-13T09:40:19
{ "title": [ 0.043029785, 0.0049476624, -0.007621765, -0.085632324, 0.020874023, -0.020111084, -0.025360107, -0.017868042, 0.010108948, 0.03692627, -0.022705078, -0.00062417984, 0.016357422, -0.032104492, -0.05114746, -0.009490967, -0.008041382, ...
http://arxiv.org/abs/2301.06916v2
Automated speech- and text-based classification of neuropsychiatric conditions in a multidiagnostic setting
Lasse Hansen, Roberta Rocca, Arndis Simonsen, Alberto Parola, Vibeke Bliksted, Nicolai Ladegaard, Dan Bang, Kristian Tylén, Ethan Weed, Søren Dinesen Østergaard, Riccardo Fusaroli
cs.CL, cs.LG, cs.SD, stat.AP
Speech patterns have been identified as potential diagnostic markers for neuropsychiatric conditions. However, most studies only compare a single clinical group to healthy controls, whereas clinical practice often requires differentiating between multiple potential diagnoses (multiclass settings). To address this, we a...
http://arxiv.org/pdf/2301.06916v2
2023-01-31T13:33:03
2023-01-13T08:24:21
{ "title": [ 0.0019779205, 0.003929138, -0.030563354, -0.07244873, 0.013137817, -0.012321472, -0.046295166, -0.008415222, 0.004096985, 0.03161621, -0.025634766, -0.0027484894, -0.00440979, -0.021026611, -0.04776001, -0.005680084, -0.010093689, 0....
http://arxiv.org/abs/2301.05402v1
In BLOOM: Creativity and Affinity in Artificial Lyrics and Art
Evan Crothers, Herna Viktor, Nathalie Japkowicz
cs.CL, cs.LG
We apply a large multilingual language model (BLOOM-176B) in open-ended generation of Chinese song lyrics, and evaluate the resulting lyrics for coherence and creativity using human reviewers. We find that current computational metrics for evaluating large language model outputs (MAUVE) have limitations in evaluation o...
http://arxiv.org/pdf/2301.05402v1
2023-01-13T06:22:22
2023-01-13T06:22:22
{ "title": [ 0.01259613, -0.004016876, -0.025283813, -0.025634766, -0.0026798248, -0.019042969, 0.023529053, -0.0015268326, -0.024871826, 0.031707764, -0.02758789, 0.005104065, 0.005680084, -0.02406311, -0.03262329, -0.030960083, -0.00093364716, ...
http://arxiv.org/abs/2301.05380v2
Prompting Neural Machine Translation with Translation Memories
Abudurexiti Reheman, Tao Zhou, Yingfeng Luo, Di Yang, Tong Xiao, Jingbo Zhu
cs.CL
Improving machine translation (MT) systems with translation memories (TMs) is of great interest to practitioners in the MT community. However, previous approaches require either a significant update of the model architecture and/or additional training efforts to make the models well-behaved when TMs are taken as additi...
http://arxiv.org/pdf/2301.05380v2
2023-02-07T08:10:07
2023-01-13T03:33:26
{ "title": [ -0.023971558, -0.01322937, -0.0046081543, -0.051330566, -0.01612854, 0.01828003, -0.039611816, 0.0022888184, -0.0012054443, 0.040039062, -0.014678955, -0.00016975403, 0.012718201, -0.016479492, -0.047821045, 0.022064209, 0.009994507, ...
http://arxiv.org/abs/2301.05327v1
Blind Judgement: Agent-Based Supreme Court Modelling With GPT
Sil Hamilton
cs.CL
We present a novel Transformer-based multi-agent system for simulating the judicial rulings of the 2010-2016 Supreme Court of the United States. We train nine separate models with the respective authored opinions of each supreme justice active ca. 2015 and test the resulting system on 96 real-world cases. We find our s...
http://arxiv.org/pdf/2301.05327v1
2023-01-12T23:07:55
2023-01-12T23:07:55
{ "title": [ 0.0155181885, 0.010734558, 0.0007638931, -0.040771484, -0.0023994446, 0.0018434525, -0.007648468, -0.030715942, 0.028198242, 0.02331543, -0.08758545, 0.012046814, 0.023422241, 0.0031147003, -0.046142578, 0.0032215118, -0.032287598, 0...
http://arxiv.org/abs/2301.05272v1
Inaccessible Neural Language Models Could Reinvigorate Linguistic Nativism
Patrick Perrine
cs.CL, cs.AI, cs.LG
Large Language Models (LLMs) have been making big waves in the machine learning community within the past few years. The impressive scalability of LLMs due to the advent of deep learning can be seen as a continuation of empiricist lingusitic methods, as opposed to rule-based linguistic methods that are grounded in a na...
http://arxiv.org/pdf/2301.05272v1
2023-01-12T19:41:47
2023-01-12T19:41:47
{ "title": [ 0.03869629, -0.009536743, 0.0061073303, -0.0031318665, -0.046295166, -0.02784729, 0.021087646, -0.019836426, 0.036865234, 0.03878784, -0.037902832, -0.04611206, 0.025360107, -0.027679443, -0.04525757, -0.009429932, -0.0052108765, 0.0...
http://arxiv.org/abs/2301.05226v1
See, Think, Confirm: Interactive Prompting Between Vision and Language Models for Knowledge-based Visual Reasoning
Zhenfang Chen, Qinhong Zhou, Yikang Shen, Yining Hong, Hao Zhang, Chuang Gan
cs.CV, cs.AI, cs.CL, cs.LG
Large pre-trained vision and language models have demonstrated remarkable capacities for various tasks. However, solving the knowledge-based visual reasoning tasks remains challenging, which requires a model to comprehensively understand image content, connect the external world knowledge, and perform step-by-step reas...
http://arxiv.org/pdf/2301.05226v1
2023-01-12T18:59:50
2023-01-12T18:59:50
{ "title": [ -0.02772522, -0.023345947, -0.02720642, -0.058502197, -0.0158844, -0.0009937286, -0.054718018, -0.0046463013, 0.030441284, 0.0027160645, -0.05859375, -0.014213562, -0.0007696152, -0.009346008, -0.04663086, -0.021133423, 0.026107788, ...
http://arxiv.org/abs/2301.05220v1
Adversarial Adaptation for French Named Entity Recognition
Arjun Choudhry, Inder Khatri, Pankaj Gupta, Aaryan Gupta, Maxime Nicol, Marie-Jean Meurs, Dinesh Kumar Vishwakarma
cs.CL, cs.AI, cs.LG
Named Entity Recognition (NER) is the task of identifying and classifying named entities in large-scale texts into predefined classes. NER in French and other relatively limited-resource languages cannot always benefit from approaches proposed for languages like English due to a dearth of large, robust datasets. In thi...
http://arxiv.org/pdf/2301.05220v1
2023-01-12T18:58:36
2023-01-12T18:58:36
{ "title": [ -0.03640747, 0.00027918816, -0.03137207, -0.037750244, 0.008544922, 0.008483887, -0.02619934, -0.00010615587, -0.015586853, 0.027740479, -0.021316528, -0.00054454803, -0.013389587, -0.0104599, -0.0390625, -0.005809784, -0.0009851456, ...
http://arxiv.org/abs/2301.05036v1
Everyone's Voice Matters: Quantifying Annotation Disagreement Using Demographic Information
Ruyuan Wan, Jaehyung Kim, Dongyeop Kang
cs.CL, cs.AI, cs.CY
In NLP annotation, it is common to have multiple annotators label the text and then obtain the ground truth labels based on the agreement of major annotators. However, annotators are individuals with different backgrounds, and minors' opinions should not be simply ignored. As annotation tasks become subjective and topi...
http://arxiv.org/pdf/2301.05036v1
2023-01-12T14:04:53
2023-01-12T14:04:53
{ "title": [ -0.0075569153, 0.00007379055, -0.02748108, -0.05987549, -0.0011291504, -0.019454956, -0.03314209, -0.0048332214, 0.008110046, 0.02696228, -0.021835327, -0.003566742, -0.02128601, -0.023406982, -0.06085205, -0.03189087, -0.0073547363, ...
http://arxiv.org/abs/2301.04962v1
A Dataset of Kurdish (Sorani) Named Entities -- An Amendment to Kurdish-BLARK Named Entities
Sazan Salar, Hossein Hassani
cs.CL
Named Entity Recognition (NER) is one of the essential applications of Natural Language Processing (NLP). It is also an instrument that plays a significant role in many other NLP applications, such as Machine Translation (MT), Information Retrieval (IR), and Part of Speech Tagging (POST). Kurdish is an under-resourced ...
http://arxiv.org/pdf/2301.04962v1
2023-01-12T12:13:44
2023-01-12T12:13:44
{ "title": [ -0.0141067505, -0.005622864, -0.050445557, -0.027053833, -0.016967773, -0.023040771, -0.054473877, -0.0067100525, -0.042022705, 0.020462036, -0.03543091, 0.02230835, -0.0043640137, -0.022735596, 0.029769897, -0.04257202, 0.014213562, ...
http://arxiv.org/abs/2301.04907v2
Think Twice: A Human-like Two-stage Conversational Agent for Emotional Response Generation
Yushan Qian, Bo Wang, Shangzhao Ma, Wu Bin, Shuo Zhang, Dongming Zhao, Kun Huang, Yuexian Hou
cs.CL, cs.HC
Towards human-like dialogue systems, current emotional dialogue approaches jointly model emotion and semantics with a unified neural network. This strategy tends to generate safe responses due to the mutual restriction between emotion and semantics, and requires rare emotion-annotated large-scale dialogue corpus. Inspi...
http://arxiv.org/pdf/2301.04907v2
2023-06-08T12:44:35
2023-01-12T10:03:56
{ "title": [ -0.015464783, -0.022613525, -0.022567749, -0.042877197, -0.011428833, -0.01020813, -0.02079773, 0.002374649, 0.0049591064, 0.042236328, -0.018600464, 0.0045280457, 0.0047340393, -0.0143966675, -0.061920166, 0.0037326813, -0.007499695, ...
http://arxiv.org/abs/2301.04883v1
SlideVQA: A Dataset for Document Visual Question Answering on Multiple Images
Ryota Tanaka, Kyosuke Nishida, Kosuke Nishida, Taku Hasegawa, Itsumi Saito, Kuniko Saito
cs.CL, cs.CV
Visual question answering on document images that contain textual, visual, and layout information, called document VQA, has received much attention recently. Although many datasets have been proposed for developing document VQA systems, most of the existing datasets focus on understanding the content relationships with...
http://arxiv.org/pdf/2301.04883v1
2023-01-12T09:00:42
2023-01-12T09:00:42
{ "title": [ -0.0072288513, -0.006916046, -0.03048706, -0.053771973, -0.007843018, -0.011764526, -0.036499023, -0.0013608932, 0.043304443, 0.011428833, -0.055389404, 0.0067749023, -0.010559082, 0.0021839142, -0.055877686, -0.01449585, 0.011360168, ...
http://arxiv.org/abs/2301.04871v2
Learning to Memorize Entailment and Discourse Relations for Persona-Consistent Dialogues
Ruijun Chen, Jin Wang, Liang-Chih Yu, Xuejie Zhang
cs.CL, cs.AI
Maintaining engagement and consistency is particularly important in dialogue systems. Existing works have improved the performance of dialogue systems by intentionally learning interlocutor personas with sophisticated network structures. One issue with this approach is that it requires more personal corpora with annota...
http://arxiv.org/pdf/2301.04871v2
2023-02-25T07:41:41
2023-01-12T08:37:00
{ "title": [ 0.0037250519, -0.00699234, -0.039154053, -0.08331299, -0.0093688965, 0.0053710938, -0.060394287, 0.025863647, -0.009628296, 0.03451538, -0.0021095276, 0.011062622, 0.0035915375, -0.013534546, -0.05355835, -0.0072669983, -0.026168823, ...
http://arxiv.org/abs/2301.08130v2
A Cohesive Distillation Architecture for Neural Language Models
Jan Philip Wahle
cs.CL
A recent trend in Natural Language Processing is the exponential growth in Language Model (LM) size, which prevents research groups without a necessary hardware infrastructure from participating in the development process. This study investigates methods for Knowledge Distillation (KD) to provide efficient alternatives...
http://arxiv.org/pdf/2301.08130v2
2023-01-29T09:39:20
2023-01-12T08:01:53
{ "title": [ -0.01386261, 0.00868988, -0.03945923, -0.072631836, 0.017669678, 0.0075416565, -0.06530762, 0.003921509, 0.012519836, 0.047180176, -0.017318726, -0.021911621, 0.0131073, -0.009689331, -0.058288574, 0.006526947, -0.011062622, 0.031036...
http://arxiv.org/abs/2301.04856v1
Multimodal Deep Learning
Cem Akkus, Luyang Chu, Vladana Djakovic, Steffen Jauch-Walser, Philipp Koch, Giacomo Loss, Christopher Marquardt, Marco Moldovan, Nadja Sauter, Maximilian Schneider, Rickmer Schulte, Karol Urbanczyk, Jann Goschenhofer, Christian Heumann, Rasmus Hvingelby, Daniel Schalk, Matthias Aßenmacher
cs.CL, cs.LG, stat.ML
This book is the result of a seminar in which we reviewed multimodal approaches and attempted to create a solid overview of the field, starting with the current state-of-the-art approaches in the two subfields of Deep Learning individually. Further, modeling frameworks are discussed where one modality is transformed in...
http://arxiv.org/pdf/2301.04856v1
2023-01-12T07:42:36
2023-01-12T07:42:36
{ "title": [ -0.02357483, 0.013656616, -0.057861328, -0.060302734, 0.019577026, -0.017349243, -0.0357666, -0.007247925, 0.004386902, 0.029281616, -0.0077705383, 0.00019943714, -0.00025749207, -0.032226562, -0.01826477, 0.015060425, -0.0011434555, ...
http://arxiv.org/abs/2301.04788v1
Language Cognition and Language Computation -- Human and Machine Language Understanding
Shaonan Wang, Nai Ding, Nan Lin, Jiajun Zhang, Chengqing Zong
cs.CL
Language understanding is a key scientific issue in the fields of cognitive and computer science. However, the two disciplines differ substantially in the specific research questions. Cognitive science focuses on analyzing the specific mechanism of the brain and investigating the brain's response to language; few studi...
http://arxiv.org/pdf/2301.04788v1
2023-01-12T02:37:00
2023-01-12T02:37:00
{ "title": [ -0.04058838, -0.01939392, -0.026046753, -0.05834961, 0.0050086975, 0.00566864, -0.019866943, 0.012145996, 0.0015516281, 0.023239136, -0.025466919, -0.0015716553, 0.0050697327, -0.025604248, -0.040863037, 0.008262634, -0.009086609, 0....
http://arxiv.org/abs/2301.04770v1
KAER: A Knowledge Augmented Pre-Trained Language Model for Entity Resolution
Liri Fang, Lan Li, Yiren Liu, Vetle I. Torvik, Bertram Ludäscher
cs.CL, cs.DB, cs.LG
Entity resolution has been an essential and well-studied task in data cleaning research for decades. Existing work has discussed the feasibility of utilizing pre-trained language models to perform entity resolution and achieved promising results. However, few works have discussed injecting domain knowledge to improve t...
http://arxiv.org/pdf/2301.04770v1
2023-01-12T00:15:40
2023-01-12T00:15:40
{ "title": [ -0.010437012, -0.013221741, -0.04562378, -0.106933594, 0.019378662, 0.002626419, -0.073791504, 0.016235352, 0.0060653687, 0.034332275, 0.0074005127, -0.010848999, 0.060058594, -0.02003479, -0.023391724, 0.028518677, -0.0072021484, 0....
http://arxiv.org/abs/2301.04761v2
NarrowBERT: Accelerating Masked Language Model Pretraining and Inference
Haoxin Li, Phillip Keung, Daniel Cheng, Jungo Kasai, Noah A. Smith
cs.CL, cs.LG
Large-scale language model pretraining is a very successful form of self-supervised learning in natural language processing, but it is increasingly expensive to perform as the models and pretraining corpora have become larger over time. We propose NarrowBERT, a modified transformer encoder that increases the throughput...
http://arxiv.org/pdf/2301.04761v2
2023-06-05T23:47:43
2023-01-11T23:45:50
{ "title": [ -0.039367676, -0.015068054, -0.06512451, -0.059631348, -0.004737854, -0.00315094, -0.068115234, 0.015113831, -0.019165039, 0.021987915, -0.007709503, 0.0090408325, 0.029144287, -0.034240723, -0.0029029846, 0.021270752, -0.009063721, ...
http://arxiv.org/abs/2301.04752v1
Semantic Web Enabled Geographic Question Answering Framework: GeoTR
Ceren Ocal Tasar, Murat Komesli, Murat Osman Unalir
cs.CL, cs.AI
With the considerable growth of linked data, researchers have focused on how to increase the availability of semantic web technologies to provide practical usages for real life systems. Question answering systems are an example of real-life systems that communicate directly with end users, understand user intention and...
http://arxiv.org/pdf/2301.04752v1
2023-01-11T23:20:43
2023-01-11T23:20:43
{ "title": [ 0.005050659, -0.010543823, -0.042785645, -0.07720947, 0.0048217773, -0.0000064373016, -0.030899048, 0.0141067505, 0.0022773743, 0.039398193, -0.046325684, -0.029006958, 0.032836914, -0.02720642, -0.032562256, -0.020248413, 0.011962891, ...
http://arxiv.org/abs/2301.04704v1
SensePOLAR: Word sense aware interpretability for pre-trained contextual word embeddings
Jan Engler, Sandipan Sikdar, Marlene Lutz, Markus Strohmaier
cs.CL, cs.LG
Adding interpretability to word embeddings represents an area of active research in text representation. Recent work has explored thepotential of embedding words via so-called polar dimensions (e.g. good vs. bad, correct vs. wrong). Examples of such recent approaches include SemAxis, POLAR, FrameAxis, and BiImp. Althou...
http://arxiv.org/pdf/2301.04704v1
2023-01-11T20:25:53
2023-01-11T20:25:53
{ "title": [ -0.0077171326, -0.0050849915, -0.032470703, -0.059448242, 0.005290985, 0.0045928955, -0.04257202, -0.011779785, 0.024978638, 0.01637268, -0.029373169, 0.0118637085, 0.007827759, -0.0063972473, -0.03439331, 0.007774353, -0.013175964, ...
http://arxiv.org/abs/2301.04647v4
EXIF as Language: Learning Cross-Modal Associations Between Images and Camera Metadata
Chenhao Zheng, Ayush Shrivastava, Andrew Owens
cs.CV, cs.CL
We learn a visual representation that captures information about the camera that recorded a given photo. To do this, we train a multimodal embedding between image patches and the EXIF metadata that cameras automatically insert into image files. Our model represents this metadata by simply converting it to text and then...
http://arxiv.org/pdf/2301.04647v4
2023-06-17T12:46:01
2023-01-11T18:59:16
{ "title": [ -0.02027893, 0.012901306, -0.04107666, -0.018737793, 0.014266968, -0.03326416, -0.013916016, -0.026748657, 0.019485474, 0.03237915, -0.05871582, 0.01058197, 0.0018472672, -0.003955841, -0.030441284, -0.00598526, -0.0052757263, 0.0531...
http://arxiv.org/abs/2301.04643v2
tieval: An Evaluation Framework for Temporal Information Extraction Systems
Hugo Sousa, Alípio Jorge, Ricardo Campos
cs.CL
Temporal information extraction (TIE) has attracted a great deal of interest over the last two decades, leading to the development of a significant number of datasets. Despite its benefits, having access to a large volume of corpora makes it difficult when it comes to benchmark TIE systems. On the one hand, different d...
http://arxiv.org/pdf/2301.04643v2
2023-04-21T15:24:09
2023-01-11T18:55:22
{ "title": [ -0.017486572, -0.0048179626, -0.06341553, -0.08282471, 0.027420044, -0.013000488, -0.037628174, 0.00907135, 0.015853882, 0.03302002, -0.032836914, 0.011932373, -0.021469116, -0.0015459061, -0.0793457, -0.013168335, -0.028015137, 0.03...
http://arxiv.org/abs/2301.04606v1
Modelling low-resource accents without accent-specific TTS frontend
Georgi Tinchev, Marta Czarnowska, Kamil Deja, Kayoko Yanagisawa, Marius Cotescu
eess.AS, cs.CL, cs.SD
This work focuses on modelling a speaker's accent that does not have a dedicated text-to-speech (TTS) frontend, including a grapheme-to-phoneme (G2P) module. Prior work on modelling accents assumes a phonetic transcription is available for the target accent, which might not be the case for low-resource, regional accent...
http://arxiv.org/pdf/2301.04606v1
2023-01-11T18:00:29
2023-01-11T18:00:29
{ "title": [ 0.010192871, 0.001045227, -0.01902771, -0.058563232, -0.0057907104, -0.0017623901, -0.042053223, 0.0070991516, 0.0035305023, 0.0579834, 0.0024585724, -0.026779175, 0.048187256, -0.036376953, -0.051574707, 0.021636963, 0.00016510487, ...
http://arxiv.org/abs/2301.04571v2
Analyzing And Improving Neural Speaker Embeddings for ASR
Christoph Lüscher, Jingjing Xu, Mohammad Zeineldeen, Ralf Schlüter, Hermann Ney
cs.CL, eess.AS, stat.ML
Neural speaker embeddings encode the speaker's speech characteristics through a DNN model and are prevalent for speaker verification tasks. However, few studies have investigated the usage of neural speaker embeddings for an ASR system. In this work, we present our efforts w.r.t integrating neural speaker embeddings in...
http://arxiv.org/pdf/2301.04571v2
2023-09-20T07:43:13
2023-01-11T16:56:03
{ "title": [ 0.004020691, 0.011459351, -0.031707764, -0.061401367, 0.015541077, -0.00093984604, -0.03201294, 0.015899658, -0.006061554, 0.039733887, -0.0065841675, -0.0027637482, 0.01537323, -0.016983032, -0.036834717, 0.0071792603, -0.008239746, ...
http://arxiv.org/abs/2301.04558v2
Learning to Exploit Temporal Structure for Biomedical Vision-Language Processing
Shruthi Bannur, Stephanie Hyland, Qianchu Liu, Fernando Pérez-García, Maximilian Ilse, Daniel C. Castro, Benedikt Boecking, Harshita Sharma, Kenza Bouzid, Anja Thieme, Anton Schwaighofer, Maria Wetscherek, Matthew P. Lungren, Aditya Nori, Javier Alvarez-Valle, Ozan Oktay
cs.CV, cs.CL
Self-supervised learning in vision-language processing exploits semantic alignment between imaging and text modalities. Prior work in biomedical VLP has mostly relied on the alignment of single image and report pairs even though clinical notes commonly refer to prior images. This does not only introduce poor alignment ...
http://arxiv.org/pdf/2301.04558v2
2023-03-16T17:12:03
2023-01-11T16:35:33
{ "title": [ -0.010398865, -0.0013446808, -0.042419434, -0.06616211, 0.010116577, -0.010620117, -0.04019165, 0.011398315, 0.02330017, 0.026519775, -0.016708374, -0.02027893, -0.004295349, -0.022003174, -0.06341553, -0.0070724487, 0.017593384, 0.0...
http://arxiv.org/abs/2301.04535v2
Few-shot Learning for Cross-Target Stance Detection by Aggregating Multimodal Embeddings
Parisa Jamadi Khiabani, Arkaitz Zubiaga
cs.CL, cs.SI
Despite the increasing popularity of the stance detection task, existing approaches are predominantly limited to using the textual content of social media posts for the classification, overlooking the social nature of the task. The stance detection task becomes particularly challenging in cross-target classification sc...
http://arxiv.org/pdf/2301.04535v2
2023-03-31T12:39:09
2023-01-11T15:52:55
{ "title": [ 0.003955841, 0.008148193, -0.03186035, -0.038024902, -0.0022583008, -0.022964478, -0.016723633, -0.010917664, 0.01828003, 0.019210815, -0.04434204, 0.018737793, -0.025848389, -0.009773254, -0.0340271, -0.01977539, -0.0008940697, 0.03...
http://arxiv.org/abs/2301.04528v1
The Role of Interactive Visualization in Explaining (Large) NLP Models: from Data to Inference
Richard Brath, Daniel Keim, Johannes Knittel, Shimei Pan, Pia Sommerauer, Hendrik Strobelt
cs.CL, cs.HC
With a constant increase of learned parameters, modern neural language models become increasingly more powerful. Yet, explaining these complex model's behavior remains a widely unsolved problem. In this paper, we discuss the role interactive visualization can play in explaining NLP models (XNLP). We motivate the use of...
http://arxiv.org/pdf/2301.04528v1
2023-01-11T15:46:52
2023-01-11T15:46:52
{ "title": [ -0.013702393, -0.015342712, -0.015716553, -0.05480957, -0.029876709, 0.008560181, -0.03491211, -0.01675415, -0.008003235, -0.0053977966, -0.0362854, -0.002532959, -0.0259552, -0.013671875, -0.028457642, -0.028900146, -0.010009766, 0....
http://arxiv.org/abs/2301.04521v1
Deteksi Depresi dan Kecemasan Pengguna Twitter Menggunakan Bidirectional LSTM
Kuncahyo Setyo Nugroho, Ismail Akbar, Affi Nizar Suksmawati, Istiadi
cs.CL
The most common mental disorders experienced by a person in daily life are depression and anxiety. Social stigma makes people with depression and anxiety neglected by their surroundings. Therefore, they turn to social media like Twitter for support. Detecting users with potential depression and anxiety disorders throug...
http://arxiv.org/pdf/2301.04521v1
2023-01-11T15:37:48
2023-01-11T15:37:48
{ "title": [ 0.03591919, 0.007331848, -0.005672455, -0.061462402, 0.011878967, -0.0015087128, -0.027572632, -0.016571045, -0.009246826, 0.054138184, -0.01687622, -0.0099487305, 0.023117065, -0.028030396, -0.04949951, -0.01701355, -0.0017852783, 0...
http://arxiv.org/abs/2301.04449v1
Diving Deep into Modes of Fact Hallucinations in Dialogue Systems
Souvik Das, Sougata Saha, Rohini K. Srihari
cs.CL, cs.AI
Knowledge Graph(KG) grounded conversations often use large pre-trained models and usually suffer from fact hallucination. Frequently entities with no references in knowledge sources and conversation history are introduced into responses, thus hindering the flow of the conversation -- existing work attempt to overcome t...
http://arxiv.org/pdf/2301.04449v1
2023-01-11T13:08:57
2023-01-11T13:08:57
{ "title": [ 0.024658203, -0.034973145, -0.007171631, -0.037231445, -0.036590576, -0.020599365, 0.029434204, 0.018920898, 0.00092983246, 0.025115967, -0.022994995, 0.0128479, 0.019363403, -0.04626465, 0.002664566, -0.0041770935, -0.00944519, 0.03...
http://arxiv.org/abs/2301.04434v1
Multilingual Entity and Relation Extraction from Unified to Language-specific Training
Zixiang Wang, Jian Yang, Tongliang Li, Jiaheng Liu, Ying Mo, Jiaqi Bai, Longtao He, Zhoujun Li
cs.CL
Entity and relation extraction is a key task in information extraction, where the output can be used for downstream NLP tasks. Existing approaches for entity and relation extraction tasks mainly focus on the English corpora and ignore other languages. Thus, it is critical to improving performance in a multilingual sett...
http://arxiv.org/pdf/2301.04434v1
2023-01-11T12:26:53
2023-01-11T12:26:53
{ "title": [ -0.014030457, -0.0053482056, -0.044769287, -0.089660645, 0.01083374, 0.0025367737, -0.057861328, 0.009002686, 0.00522995, 0.029251099, 0.013214111, 0.011207581, 0.016204834, -0.027526855, -0.010955811, 0.024719238, -0.020568848, 0.04...
http://arxiv.org/abs/2301.04408v1
GPT as Knowledge Worker: A Zero-Shot Evaluation of (AI)CPA Capabilities
Jillian Bommarito, Michael Bommarito, Daniel Martin Katz, Jessica Katz
cs.CL, cs.AI, cs.CY
The global economy is increasingly dependent on knowledge workers to meet the needs of public and private organizations. While there is no single definition of knowledge work, organizations and industry groups still attempt to measure individuals' capability to engage in it. The most comprehensive assessment of capabil...
http://arxiv.org/pdf/2301.04408v1
2023-01-11T11:30:42
2023-01-11T11:30:42
{ "title": [ -0.02671814, -0.009536743, -0.041900635, -0.015457153, -0.020996094, -0.01725769, -0.007926941, -0.013038635, 0.0036010742, 0.020629883, -0.07128906, 0.003145218, -0.022262573, -0.0023651123, -0.050750732, -0.038116455, -0.0074768066, ...
http://arxiv.org/abs/2301.04388v3
Perceive and predict: self-supervised speech representation based loss functions for speech enhancement
George Close, William Ravenscroft, Thomas Hain, Stefan Goetze
cs.SD, cs.CL, cs.LG, eess.AS
Recent work in the domain of speech enhancement has explored the use of self-supervised speech representations to aid in the training of neural speech enhancement models. However, much of this work focuses on using the deepest or final outputs of self supervised speech representation models, rather than the earlier fea...
http://arxiv.org/pdf/2301.04388v3
2023-06-26T09:31:53
2023-01-11T10:20:56
{ "title": [ 0.0026741028, 0.011390686, -0.045318604, -0.06311035, 0.017150879, -0.0008177757, -0.019973755, -0.0035572052, -0.012435913, 0.047058105, -0.0029773712, -0.004486084, 0.017669678, -0.03036499, -0.04147339, 0.006134033, -0.0052986145, ...
http://arxiv.org/abs/2301.04366v1
Multimodal Inverse Cloze Task for Knowledge-based Visual Question Answering
Paul Lerner, Olivier Ferret, Camille Guinaudeau
cs.CL, cs.IR, cs.LG, cs.MM
We present a new pre-training method, Multimodal Inverse Cloze Task, for Knowledge-based Visual Question Answering about named Entities (KVQAE). KVQAE is a recently introduced task that consists in answering questions about named entities grounded in a visual context using a Knowledge Base. Therefore, the interaction b...
http://arxiv.org/pdf/2301.04366v1
2023-01-11T09:16:34
2023-01-11T09:16:34
{ "title": [ -0.033569336, -0.003501892, -0.038116455, -0.030914307, -0.02470398, -0.0069084167, -0.03439331, 0.0040245056, 0.029846191, 0.03778076, -0.014808655, -0.0053863525, -0.002521515, -0.003200531, -0.07055664, -0.00011301041, 0.010116577, ...
http://arxiv.org/abs/2301.04347v3
Counteracts: Testing Stereotypical Representation in Pre-trained Language Models
Damin Zhang, Julia Rayz, Romila Pradhan
cs.CL
Recently, language models have demonstrated strong performance on various natural language understanding tasks. Language models trained on large human-generated corpus encode not only a significant amount of human knowledge, but also the human stereotype. As more and more downstream tasks have integrated language model...
http://arxiv.org/pdf/2301.04347v3
2023-04-07T17:12:23
2023-01-11T07:52:59
{ "title": [ 0.006061554, -0.0051879883, -0.016357422, -0.05621338, -0.0009431839, -0.013343811, -0.029647827, -0.012290955, -0.007167816, 0.026611328, -0.01210022, 0.0059547424, 0.0045661926, -0.023956299, -0.036376953, -0.0003039837, -0.015045166, ...
http://arxiv.org/abs/2301.04339v1
Topics in Contextualised Attention Embeddings
Mozhgan Talebpour, Alba Garcia Seco de Herrera, Shoaib Jameel
cs.CL, cs.IR
Contextualised word vectors obtained via pre-trained language models encode a variety of knowledge that has already been exploited in applications. Complementary to these language models are probabilistic topic models that learn thematic patterns from the text. Recent work has demonstrated that conducting clustering on...
http://arxiv.org/pdf/2301.04339v1
2023-01-11T07:26:19
2023-01-11T07:26:19
{ "title": [ 0.008285522, 0.020431519, -0.038360596, -0.06109619, 0.020889282, -0.010543823, -0.03173828, -0.025909424, 0.04058838, 0.025054932, -0.011413574, 0.012519836, 0.005935669, -0.005695343, -0.014457703, 0.026123047, -0.0075531006, 0.024...
http://arxiv.org/abs/2301.04327v1
Dual Learning for Large Vocabulary On-Device ASR
Cal Peyser, Ronny Huang, Tara Sainath, Rohit Prabhavalkar, Michael Picheny, Kyunghyun Cho
cs.CL, cs.AI
Dual learning is a paradigm for semi-supervised machine learning that seeks to leverage unsupervised data by solving two opposite tasks at once. In this scheme, each model is used to generate pseudo-labels for unlabeled examples that are used to train the other model. Dual learning has seen some use in speech processin...
http://arxiv.org/pdf/2301.04327v1
2023-01-11T06:32:28
2023-01-11T06:32:28
{ "title": [ -0.02142334, -0.017333984, -0.036468506, -0.06185913, -0.009712219, -0.0047569275, -0.02949524, 0.036895752, -0.051635742, 0.036743164, -0.026870728, 0.015525818, -0.011787415, -0.019683838, -0.005367279, -0.020935059, 0.0027637482, ...
http://arxiv.org/abs/2301.04312v4
Word-Graph2vec: An efficient word embedding approach on word co-occurrence graph using random walk sampling
Wenting Li, Jiahong Xue, Xi Zhang, Huacan Chen, Zeyu Chen, Yuanzhe Cai
cs.CL
Word embedding has become ubiquitous and is widely used in various text mining and natural language processing (NLP) tasks, such as information retrieval, semantic analysis, and machine translation, among many others. Unfortunately, it is prohibitively expensive to train the word embedding in a relatively large corpus....
http://arxiv.org/pdf/2301.04312v4
2023-05-14T09:35:34
2023-01-11T05:21:00
{ "title": [ -0.017028809, -0.010055542, -0.03466797, -0.043823242, -0.0012264252, 0.002336502, -0.031555176, -0.014373779, 0.012496948, 0.024597168, -0.023239136, 0.044311523, -0.017532349, -0.011627197, -0.019638062, 0.010421753, -0.021835327, ...
http://arxiv.org/abs/2301.04283v2
MGeo: Multi-Modal Geographic Pre-Training Method
Ruixue Ding, Boli Chen, Pengjun Xie, Fei Huang, Xin Li, Qiang Zhang, Yao Xu
cs.CL
As a core task in location-based services (LBS) (e.g., navigation maps), query and point of interest (POI) matching connects users' intent with real-world geographic information. Recently, pre-trained models (PTMs) have made advancements in many natural language processing (NLP) tasks. Generic text-based PTMs do not ha...
http://arxiv.org/pdf/2301.04283v2
2023-05-24T04:20:27
2023-01-11T03:05:12
{ "title": [ 0.0015420914, -0.0059432983, -0.025863647, -0.023742676, -0.011680603, -0.025817871, -0.014923096, -0.03503418, -0.029556274, 0.03451538, -0.006839752, 0.0006289482, 0.030776978, -0.048858643, 0.0067443848, -0.0075912476, 0.005996704, ...
http://arxiv.org/abs/2301.04253v2
Towards Answering Climate Questionnaires from Unstructured Climate Reports
Daniel Spokoyny, Tanmay Laud, Tom Corringham, Taylor Berg-Kirkpatrick
cs.CL, cs.AI, cs.IR, cs.LG
The topic of Climate Change (CC) has received limited attention in NLP despite its urgency. Activists and policymakers need NLP tools to effectively process the vast and rapidly growing unstructured textual climate reports into structured form. To tackle this challenge we introduce two new large-scale climate questionn...
http://arxiv.org/pdf/2301.04253v2
2023-07-27T18:13:40
2023-01-11T00:22:56
{ "title": [ -0.032196045, -0.0020179749, -0.05316162, -0.048736572, 0.0020179749, -0.009925842, -0.008201599, -0.0026760101, 0.003250122, 0.017807007, -0.023376465, -0.010223389, -0.058898926, -0.021408081, -0.06124878, -0.018112183, -0.0035209656, ...
http://arxiv.org/abs/2301.04248v1
Predicting Hateful Discussions on Reddit using Graph Transformer Networks and Communal Context
Liam Hebert, Lukasz Golab, Robin Cohen
cs.CL, cs.LG, cs.SI
We propose a system to predict harmful discussions on social media platforms. Our solution uses contextual deep language models and proposes the novel idea of integrating state-of-the-art Graph Transformer Networks to analyze all conversations that follow an initial post. This framework also supports adapting to future...
http://arxiv.org/pdf/2301.04248v1
2023-01-10T23:47:13
2023-01-10T23:47:13
{ "title": [ 0.023422241, 0.008354187, -0.029556274, -0.034820557, 0.013801575, -0.025054932, -0.008575439, -0.01826477, -0.022415161, 0.036346436, -0.007873535, 0.027145386, 0.026809692, -0.013435364, -0.019302368, 0.0058059692, -0.004009247, 0....
http://arxiv.org/abs/2301.04230v1
User-Centered Security in Natural Language Processing
Chris Emmery
cs.CL, cs.CY
This dissertation proposes a framework of user-centered security in Natural Language Processing (NLP), and demonstrates how it can improve the accessibility of related research. Accordingly, it focuses on two security domains within NLP with great public interest. First, that of author profiling, which can be employed ...
http://arxiv.org/pdf/2301.04230v1
2023-01-10T22:34:19
2023-01-10T22:34:19
{ "title": [ -0.007411957, -0.0010614395, -0.04083252, -0.053833008, 0.009605408, -0.014556885, -0.014717102, 0.0209198, -0.04171753, 0.024780273, -0.023086548, -0.0030899048, 0.0037117004, -0.020858765, -0.03704834, -0.01424408, -0.026123047, 0....
http://arxiv.org/abs/2301.04213v2
Does Localization Inform Editing? Surprising Differences in Causality-Based Localization vs. Knowledge Editing in Language Models
Peter Hase, Mohit Bansal, Been Kim, Asma Ghandeharioun
cs.LG, cs.AI, cs.CL
Language models learn a great quantity of factual information during pretraining, and recent work localizes this information to specific model weights like mid-layer MLP weights. In this paper, we find that we can change how a fact is stored in a model by editing weights that are in a different location than where exis...
http://arxiv.org/pdf/2301.04213v2
2023-10-16T17:42:58
2023-01-10T21:26:08
{ "title": [ -0.0541687, 0.00233078, -0.035125732, -0.031188965, -0.032104492, -0.006008148, -0.06591797, -0.019714355, -0.02822876, 0.022476196, -0.0066337585, -0.004562378, 0.033111572, -0.037902832, -0.006580353, 0.0024471283, -0.022094727, 0....
http://arxiv.org/abs/2301.04110v1
Structured Case-based Reasoning for Inference-time Adaptation of Text-to-SQL parsers
Abhijeet Awasthi, Soumen Chakrabarti, Sunita Sarawagi
cs.CL, cs.AI
Inference-time adaptation methods for semantic parsing are useful for leveraging examples from newly-observed domains without repeated fine-tuning. Existing approaches typically bias the decoder by simply concatenating input-output example pairs (cases) from the new domain at the encoder's input in a Seq-to-Seq model. ...
http://arxiv.org/pdf/2301.04110v1
2023-01-10T18:20:42
2023-01-10T18:20:42
{ "title": [ -0.006450653, -0.023330688, -0.039093018, -0.11462402, 0.013168335, 0.013885498, -0.06585693, 0.03829956, 0.012939453, 0.016067505, -0.049041748, 0.0054016113, -0.009864807, -0.011543274, -0.028198242, -0.037353516, -0.009475708, 0.0...
http://arxiv.org/abs/2301.04098v2
Investigating Conversational Search Behavior For Domain Exploration
Phillip Schneider, Anum Afzal, Juraj Vladika, Daniel Braun, Florian Matthes
cs.CL
Conversational search has evolved as a new information retrieval paradigm, marking a shift from traditional search systems towards interactive dialogues with intelligent search agents. This change especially affects exploratory information-seeking contexts, where conversational search systems can guide the discovery of...
http://arxiv.org/pdf/2301.04098v2
2023-02-27T15:49:16
2023-01-10T17:43:03
{ "title": [ 0.0021743774, -0.009628296, -0.01751709, -0.016479492, 0.0025234222, -0.02609253, 0.0020217896, -0.020492554, 0.008010864, 0.02519226, 0.0011577606, -0.0044136047, -0.013702393, -0.023345947, -0.04876709, -0.00374794, -0.033569336, 0...
http://arxiv.org/abs/2301.08155v1
AI Insights into Theoretical Physics and the Swampland Program: A Journey Through the Cosmos with ChatGPT
Kay Lehnert
physics.pop-ph, cs.AI, cs.CL
In this case study, we explore the capabilities and limitations of ChatGPT, a natural language processing model developed by OpenAI, in the field of string theoretical swampland conjectures. We find that it is effective at paraphrasing and explaining concepts in a variety of styles, but not at genuinely connecting conc...
http://arxiv.org/pdf/2301.08155v1
2023-01-10T16:57:16
2023-01-10T16:57:16
{ "title": [ -0.023330688, -0.0118255615, -0.042663574, -0.021759033, -0.016021729, -0.03048706, 0.0137786865, 0.010482788, -0.015213013, 0.034606934, 0.00422287, 0.03326416, 0.018341064, -0.04321289, 0.013191223, 0.0071258545, 0.0146484375, 0.01...
http://arxiv.org/abs/2301.04013v1
There is No Big Brother or Small Brother: Knowledge Infusion in Language Models for Link Prediction and Question Answering
Ankush Agarwal, Sakharam Gawade, Sachin Channabasavarajendra, Pushpak Bhattacharyya
cs.CL, cs.AI, cs.LG
The integration of knowledge graphs with deep learning is thriving in improving the performance of various natural language processing (NLP) tasks. In this paper, we focus on knowledge-infused link prediction and question answering using language models, T5, and BLOOM across three domains: Aviation, Movie, and Web. In ...
http://arxiv.org/pdf/2301.04013v1
2023-01-10T14:59:33
2023-01-10T14:59:33
{ "title": [ -0.0064086914, -0.020019531, -0.03781128, -0.06750488, -0.014816284, -0.0036201477, -0.05557251, 0.012580872, 0.007194519, 0.03866577, -0.0033092499, 0.010734558, 0.018203735, -0.016464233, -0.04623413, 0.016036987, 0.003862381, 0.03...
http://arxiv.org/abs/2301.03980v1
Language Models sounds the Death Knell of Knowledge Graphs
Kunal Suri, Atul Singh, Prakhar Mishra, Swapna Sourav Rout, Rajesh Sabapathy
cs.CL, cs.AI
Healthcare domain generates a lot of unstructured and semi-structured text. Natural Language processing (NLP) has been used extensively to process this data. Deep Learning based NLP especially Large Language Models (LLMs) such as BERT have found broad acceptance and are used extensively for many applications. A Languag...
http://arxiv.org/pdf/2301.03980v1
2023-01-10T14:20:15
2023-01-10T14:20:15
{ "title": [ -0.042755127, -0.031585693, -0.053009033, -0.046905518, -0.0736084, -0.007183075, -0.051208496, 0.039001465, -0.0029182434, 0.04812622, -0.040496826, -0.0104904175, 0.026733398, -0.016708374, -0.030258179, -0.00299263, 0.018920898, 0...
http://arxiv.org/abs/2301.03971v1
Unsupervised Mandarin-Cantonese Machine Translation
Megan Dare, Valentina Fajardo Diaz, Averie Ho Zoen So, Yifan Wang, Shibingfeng Zhang
cs.CL
Advancements in unsupervised machine translation have enabled the development of machine translation systems that can translate between languages for which there is not an abundance of parallel data available. We explored unsupervised machine translation between Mandarin Chinese and Cantonese. Despite the vast number o...
http://arxiv.org/pdf/2301.03971v1
2023-01-10T14:09:40
2023-01-10T14:09:40
{ "title": [ -0.034454346, 0.004306793, -0.025466919, -0.022994995, -0.016159058, -0.0009293556, -0.03552246, -0.013900757, -0.009185791, 0.06036377, -0.026107788, 0.002521515, 0.047943115, -0.02696228, -0.028182983, 0.021362305, 0.001420021, 0.0...
http://arxiv.org/abs/2301.03957v1
AI based approach to Trailer Generation for Online Educational Courses
Prakhar Mishra, Chaitali Diwan, Srinath Srinivasa, G. Srinivasaraghavan
cs.CL, cs.AI
In this paper, we propose an AI based approach to Trailer Generation in the form of short videos for online educational courses. Trailers give an overview of the course to the learners and help them make an informed choice about the courses they want to learn. It also helps to generate curiosity and interest among the ...
http://arxiv.org/pdf/2301.03957v1
2023-01-10T13:33:08
2023-01-10T13:33:08
{ "title": [ -0.023269653, -0.00932312, -0.035827637, -0.030822754, -0.006465912, -0.006793976, -0.009002686, -0.00655365, -0.020309448, 0.041259766, -0.01525116, 0.013046265, 0.0073165894, -0.023468018, -0.005531311, -0.016830444, -0.01209259, 0...
http://arxiv.org/abs/2301.03953v2
Channel-aware Decoupling Network for Multi-turn Dialogue Comprehension
Zhuosheng Zhang, Hai Zhao, Longxiang Liu
cs.CL, cs.AI, cs.HC
Training machines to understand natural language and interact with humans is one of the major goals of artificial intelligence. Recent years have witnessed an evolution from matching networks to pre-trained language models (PrLMs). In contrast to the plain-text modeling as the focus of the PrLMs, dialogue texts involve...
http://arxiv.org/pdf/2301.03953v2
2023-01-11T02:21:04
2023-01-10T13:18:25
{ "title": [ -0.009048462, -0.0020656586, -0.009506226, -0.07196045, -0.021316528, 0.016448975, -0.059265137, 0.016937256, 0.020050049, 0.026489258, -0.0035533905, -0.022720337, 0.017654419, -0.041992188, -0.0395813, -0.006095886, -0.030471802, 0...
http://arxiv.org/abs/2301.03848v1
The use of new technologies to support Public Administration. Sentiment analysis and the case of the app IO
Vincenzo Miracula, Antonio Picone
cs.SI, cs.CL
App IO is an app developed for the Italian PA. It is definitely useful for citizens to interact with the PA and to get services that were not digitized yet. Nevertheless, it was not perceived in a good way by the citizens and it has been criticized. As we wanted to find the root that caused all these bad reviews we scr...
http://arxiv.org/pdf/2301.03848v1
2023-01-10T08:41:25
2023-01-10T08:41:25
{ "title": [ 0.03555298, -0.024261475, -0.018005371, -0.017593384, -0.020080566, -0.02279663, -0.0006971359, 0.0008416176, -0.011199951, 0.037200928, -0.052978516, 0.013435364, -0.0014629364, -0.012809753, -0.020065308, -0.046081543, 0.00063991547, ...
http://arxiv.org/abs/2301.03819v1
Streaming Punctuation: A Novel Punctuation Technique Leveraging Bidirectional Context for Continuous Speech Recognition
Piyush Behre, Sharman Tan, Padma Varadharajan, Shuangyu Chang
cs.CL
While speech recognition Word Error Rate (WER) has reached human parity for English, continuous speech recognition scenarios such as voice typing and meeting transcriptions still suffer from segmentation and punctuation problems, resulting from irregular pausing patterns or slow speakers. Transformer sequence tagging m...
http://arxiv.org/pdf/2301.03819v1
2023-01-10T07:07:20
2023-01-10T07:07:20
{ "title": [ 0.017333984, 0.01071167, -0.044158936, -0.087768555, 0.016677856, -0.0042419434, -0.06048584, 0.016067505, -0.03262329, 0.055236816, -0.05307007, 0.020065308, 0.0030078888, 0.0027275085, -0.046203613, -0.036499023, -0.013290405, 0.05...
http://arxiv.org/abs/2301.03801v1
UnifySpeech: A Unified Framework for Zero-shot Text-to-Speech and Voice Conversion
Haogeng Liu, Tao Wang, Ruibo Fu, Jiangyan Yi, Zhengqi Wen, Jianhua Tao
cs.SD, cs.CL, cs.LG, eess.AS
Text-to-speech (TTS) and voice conversion (VC) are two different tasks both aiming at generating high quality speaking voice according to different input modality. Due to their similarity, this paper proposes UnifySpeech, which brings TTS and VC into a unified framework for the first time. The model is based on the ass...
http://arxiv.org/pdf/2301.03801v1
2023-01-10T06:06:57
2023-01-10T06:06:57
{ "title": [ -0.0075912476, -0.010559082, -0.026794434, -0.066345215, 0.005908966, 0.0024547577, -0.03768921, 0.008377075, 0.012123108, 0.060272217, -0.0014915466, 0.006702423, 0.036010742, -0.018554688, -0.05038452, 0.02104187, -0.0051574707, 0....
http://arxiv.org/abs/2301.03765v1
Cross-Model Comparative Loss for Enhancing Neuronal Utility in Language Understanding
Yunchang Zhu, Liang Pang, Kangxi Wu, Yanyan Lan, Huawei Shen, Xueqi Cheng
cs.CL, cs.IR, cs.LG
Current natural language understanding (NLU) models have been continuously scaling up, both in terms of model size and input context, introducing more hidden and input neurons. While this generally improves performance on average, the extra neurons do not yield a consistent improvement for all instances. This is becaus...
http://arxiv.org/pdf/2301.03765v1
2023-01-10T03:04:27
2023-01-10T03:04:27
{ "title": [ -0.021987915, -0.010757446, -0.022537231, -0.023376465, -0.035003662, -0.002363205, -0.032836914, -0.015792847, -0.0021152496, 0.03173828, -0.014801025, -0.0020122528, 0.007385254, -0.016571045, -0.03640747, 0.017044067, -0.010055542, ...
http://arxiv.org/abs/2301.04589v1
Memory Augmented Large Language Models are Computationally Universal
Dale Schuurmans
cs.CL, cs.FL
We show that transformer-based large language models are computationally universal when augmented with an external memory. Any deterministic language model that conditions on strings of bounded length is equivalent to a finite automaton, hence computationally limited. However, augmenting such models with a read-write m...
http://arxiv.org/pdf/2301.04589v1
2023-01-10T02:37:44
2023-01-10T02:37:44
{ "title": [ -0.0013284683, -0.0003707409, -0.031204224, -0.07684326, -0.023742676, -0.010025024, -0.07098389, 0.016159058, 0.00092077255, 0.048614502, -0.0112838745, -0.0015974045, 0.027740479, -0.035125732, -0.021774292, 0.028076172, -0.0021705627,...
http://arxiv.org/abs/2301.03728v1
Scaling Laws for Generative Mixed-Modal Language Models
Armen Aghajanyan, Lili Yu, Alexis Conneau, Wei-Ning Hsu, Karen Hambardzumyan, Susan Zhang, Stephen Roller, Naman Goyal, Omer Levy, Luke Zettlemoyer
cs.CL, cs.AI, cs.LG
Generative language models define distributions over sequences of tokens that can represent essentially any combination of data modalities (e.g., any permutation of image tokens from VQ-VAEs, speech tokens from HuBERT, BPE tokens for language or code, and so on). To better understand the scaling properties of such mixe...
http://arxiv.org/pdf/2301.03728v1
2023-01-10T00:20:06
2023-01-10T00:20:06
{ "title": [ -0.036376953, -0.006088257, -0.045562744, -0.05126953, -0.0062789917, -0.000112235546, -0.07849121, 0.0024375916, -0.0039138794, 0.04660034, -0.014785767, -0.0014152527, 0.022613525, -0.014877319, -0.048614502, 0.0017642975, -0.012580872...
http://arxiv.org/abs/2301.03726v2
Neighborhood-Regularized Self-Training for Learning with Few Labels
Ran Xu, Yue Yu, Hejie Cui, Xuan Kan, Yanqiao Zhu, Joyce Ho, Chao Zhang, Carl Yang
cs.LG, cs.CL
Training deep neural networks (DNNs) with limited supervision has been a popular research topic as it can significantly alleviate the annotation burden. Self-training has been successfully applied in semi-supervised learning tasks, but one drawback of self-training is that it is vulnerable to the label noise from incor...
http://arxiv.org/pdf/2301.03726v2
2023-02-15T20:29:50
2023-01-10T00:07:33
{ "title": [ -0.017959595, 0.011528015, -0.03668213, -0.044952393, 0.020370483, 0.0020313263, -0.024459839, -0.014305115, -0.015808105, 0.02168274, -0.0027313232, 0.024215698, -0.03881836, -0.02571106, -0.024627686, 0.006259918, 0.016052246, 0.03...
http://arxiv.org/abs/2301.03559v1
Color Me Intrigued: Quantifying Usage of Colors in Fiction
Siyan Li
cs.CL
We present preliminary results in quantitative analyses of color usage in selected authors' works from LitBank. Using Glasgow Norms, human ratings on 5000+ words, we measure attributes of nouns dependent on color terms. Early results demonstrate a significant increase in noun concreteness over time. We also propose fut...
http://arxiv.org/pdf/2301.03559v1
2023-01-09T18:20:10
2023-01-09T18:20:10
{ "title": [ -0.015602112, 0.007888794, -0.05316162, 0.0072021484, -0.010749817, -0.038848877, 0.017700195, -0.020706177, -0.04724121, 0.035583496, -0.0021381378, 0.041229248, -0.0065345764, -0.033477783, -0.003768921, -0.013008118, -0.014045715, ...
http://arxiv.org/abs/2301.03531v1
Leveraging Contextual Relatedness to Identify Suicide Documentation in Clinical Notes through Zero Shot Learning
Terri Elizabeth Workman, Joseph L. Goulet, Cynthia A. Brandt, Allison R. Warren, Jacob Eleazer, Melissa Skanderson, Luke Lindemann, John R. Blosnich, John O Leary, Qing Zeng Treitler
cs.AI, cs.CL, I.2.7; J.3
Identifying suicidality including suicidal ideation, attempts, and risk factors in electronic health record data in clinical notes is difficult. A major difficulty is the lack of training samples given the small number of true positive instances among the increasingly large number of patients being screened. This paper...
http://arxiv.org/pdf/2301.03531v1
2023-01-09T17:26:07
2023-01-09T17:26:07
{ "title": [ 0.049591064, -0.006591797, -0.010253906, -0.083862305, 0.0006637573, -0.01411438, -0.039398193, -0.009773254, 0.0082092285, 0.056671143, -0.013420105, 0.005886078, 0.026687622, -0.025497437, -0.02998352, -0.00059604645, 0.0026931763, ...
http://arxiv.org/abs/2301.03458v1
Mining Healthcare Procurement Data Using Text Mining and Natural Language Processing -- Reflection From An Industrial Project
Ziqi Zhang, Tomas Jasaitis, Richard Freeman, Rowida Alfrjani, Adam Funk
cs.CL, cs.AI
While text mining and NLP research has been established for decades, there remain gaps in the literature that reports the use of these techniques in building real-world applications. For example, they typically look at single and sometimes simplified tasks, and do not discuss in-depth data heterogeneity and inconsisten...
http://arxiv.org/pdf/2301.03458v1
2023-01-09T15:59:55
2023-01-09T15:59:55
{ "title": [ -0.032958984, -0.0052833557, -0.06463623, -0.06359863, 0.0058021545, -0.0067863464, -0.052886963, 0.018981934, -0.050933838, 0.014762878, -0.06738281, 0.038330078, -0.04788208, -0.00026750565, -0.010276794, -0.055114746, -0.012680054, ...
http://arxiv.org/abs/2301.03447v1
Automatic Standardization of Arabic Dialects for Machine Translation
Abidrabbo Alnassan
cs.CL
Based on an annotated multimedia corpus, television series Mar{\=a}y{\=a} 2013, we dig into the question of ''automatic standardization'' of Arabic dialects for machine translation. Here we distinguish between rule-based machine translation and statistical machine translation. Machine translation from Arabic most of th...
http://arxiv.org/pdf/2301.03447v1
2023-01-09T15:52:40
2023-01-09T15:52:40
{ "title": [ -0.054382324, -0.0076179504, -0.033172607, -0.04940796, -0.012451172, 0.015304565, -0.05178833, 0.017715454, -0.018722534, 0.054656982, 0.009742737, -0.00018441677, 0.014579773, -0.040130615, -0.038208008, 0.019042969, 0.021575928, -...
http://arxiv.org/abs/2301.03416v1
ERNIE 3.0 Tiny: Frustratingly Simple Method to Improve Task-Agnostic Distillation Generalization
Weixin Liu, Xuyi Chen, Jiaxiang Liu, Shikun Feng, Yu Sun, Hao Tian, Hua Wu
cs.CL
Task-agnostic knowledge distillation attempts to address the problem of deploying large pretrained language model in resource-constrained scenarios by compressing a large pretrained model called teacher into a smaller one called student such that the student can be directly finetuned on downstream tasks and retains com...
http://arxiv.org/pdf/2301.03416v1
2023-01-09T15:12:50
2023-01-09T15:12:50
{ "title": [ 0.019042969, 0.019760132, -0.07098389, -0.06222534, 0.006298065, -0.0259552, -0.061187744, -0.0031337738, 0.021606445, 0.052581787, -0.03375244, -0.024307251, 0.04119873, -0.02758789, -0.044311523, -0.0050849915, 0.01436615, 0.000052...
http://arxiv.org/abs/2301.03391v2
AI2: The next leap toward native language based and explainable machine learning framework
Jean-Sébastien Dessureault, Daniel Massicotte
cs.LG, cs.CL
The machine learning frameworks flourished in the last decades, allowing artificial intelligence to get out of academic circles to be applied to enterprise domains. This field has significantly advanced, but there is still some meaningful improvement to reach the subsequent expectations. The proposed framework, named A...
http://arxiv.org/pdf/2301.03391v2
2023-01-12T20:53:50
2023-01-09T14:48:35
{ "title": [ -0.0501709, -0.018341064, -0.03918457, -0.050048828, -0.0008659363, 0.0066108704, -0.051513672, -0.016448975, -0.010383606, 0.018341064, -0.000415802, -0.0053977966, -0.005569458, -0.03579712, -0.026779175, 0.00894928, 0.00018906593, ...
http://arxiv.org/abs/2301.03353v2
Learning Bidirectional Action-Language Translation with Limited Supervision and Incongruent Input
Ozan Özdemir, Matthias Kerzel, Cornelius Weber, Jae Hee Lee, Muhammad Burhan Hafez, Patrick Bruns, Stefan Wermter
cs.CL, cs.AI, cs.NE, cs.RO
Human infant learning happens during exploration of the environment, by interaction with objects, and by listening to and repeating utterances casually, which is analogous to unsupervised learning. Only occasionally, a learning infant would receive a matching verbal description of an action it is committing, which is s...
http://arxiv.org/pdf/2301.03353v2
2023-02-22T15:54:47
2023-01-09T14:09:09
{ "title": [ -0.014884949, -0.009613037, -0.025024414, -0.06188965, -0.00062704086, 0.00349617, -0.045166016, 0.003730774, -0.0057868958, 0.03741455, -0.021377563, 0.018615723, -0.003414154, -0.008285522, -0.036346436, -0.017227173, -0.0033302307, ...
http://arxiv.org/abs/2301.03344v1
Universal Multimodal Representation for Language Understanding
Zhuosheng Zhang, Kehai Chen, Rui Wang, Masao Utiyama, Eiichiro Sumita, Zuchao Li, Hai Zhao
cs.CL, cs.AI, cs.CV
Representation learning is the foundation of natural language processing (NLP). This work presents new methods to employ visual information as assistant signals to general NLP tasks. For each sentence, we first retrieve a flexible number of images either from a light topic-image lookup table extracted over the existing...
http://arxiv.org/pdf/2301.03344v1
2023-01-09T13:54:11
2023-01-09T13:54:11
{ "title": [ -0.022903442, 0.0027332306, -0.028884888, -0.06756592, -0.007949829, 0.0036563873, -0.06939697, -0.0048446655, 0.02218628, 0.035064697, -0.008880615, -0.0037403107, 0.01449585, -0.017349243, -0.041748047, 0.013015747, -0.004722595, 0...
http://arxiv.org/abs/2301.03319v1
FullStop:Punctuation and Segmentation Prediction for Dutch with Transformers
Vincent Vandeghinste, Oliver Guhr
cs.CL, cs.AI, I.2.7
When applying automated speech recognition (ASR) for Belgian Dutch (Van Dyck et al. 2021), the output consists of an unsegmented stream of words, without any punctuation. A next step is to perform segmentation and insert punctuation, making the ASR output more readable and easy to manually correct. As far as we know th...
http://arxiv.org/pdf/2301.03319v1
2023-01-09T13:12:05
2023-01-09T13:12:05
{ "title": [ -0.009269714, 0.004348755, -0.0077552795, -0.08258057, -0.0050849915, 0.024108887, -0.074523926, -0.0037670135, -0.035095215, 0.028579712, -0.020446777, 0.011260986, 0.024993896, -0.019622803, -0.026138306, 0.0059165955, -0.013671875, ...
http://arxiv.org/abs/2301.03282v1
Universal Information Extraction as Unified Semantic Matching
Jie Lou, Yaojie Lu, Dai Dai, Wei Jia, Hongyu Lin, Xianpei Han, Le Sun, Hua Wu
cs.CL
The challenge of information extraction (IE) lies in the diversity of label schemas and the heterogeneity of structures. Traditional methods require task-specific model design and rely heavily on expensive supervision, making them difficult to generalize to new schemas. In this paper, we decouple IE into two basic abil...
http://arxiv.org/pdf/2301.03282v1
2023-01-09T11:51:31
2023-01-09T11:51:31
{ "title": [ -0.03338623, -0.0011587143, -0.05105591, -0.07800293, 0.014343262, 0.011009216, -0.050231934, -0.0060157776, 0.016616821, 0.04119873, -0.009384155, 0.014633179, 0.0036716461, -0.009185791, -0.04046631, 0.025115967, -0.013572693, 0.03...
http://arxiv.org/abs/2301.03252v1
Active Learning for Abstractive Text Summarization
Akim Tsvigun, Ivan Lysenko, Danila Sedashov, Ivan Lazichny, Eldar Damirov, Vladimir Karlov, Artemy Belousov, Leonid Sanochkin, Maxim Panov, Alexander Panchenko, Mikhail Burtsev, Artem Shelmanov
cs.CL
Construction of human-curated annotated datasets for abstractive text summarization (ATS) is very time-consuming and expensive because creating each instance requires a human annotator to read a long document and compose a shorter summary that would preserve the key information relayed by the original document. Active ...
http://arxiv.org/pdf/2301.03252v1
2023-01-09T10:33:14
2023-01-09T10:33:14
{ "title": [ -0.008483887, 0.0064964294, -0.00566864, -0.06542969, -0.0032787323, 0.011390686, -0.034698486, -0.017318726, 0.02645874, 0.05331421, -0.011070251, 0.025924683, -0.023498535, -0.005809784, -0.060760498, 0.01436615, -0.0020179749, 0.0...
http://arxiv.org/abs/2301.03238v1
MAQA: A Multimodal QA Benchmark for Negation
Judith Yue Li, Aren Jansen, Qingqing Huang, Joonseok Lee, Ravi Ganti, Dima Kuzmin
cs.CL, cs.AI, cs.LG, cs.SD, eess.AS
Multimodal learning can benefit from the representation power of pretrained Large Language Models (LLMs). However, state-of-the-art transformer based LLMs often ignore negations in natural language and there is no existing benchmark to quantitatively evaluate whether multimodal transformers inherit this weakness. In th...
http://arxiv.org/pdf/2301.03238v1
2023-01-09T10:11:23
2023-01-09T10:11:23
{ "title": [ -0.0065078735, -0.013931274, -0.019042969, -0.055511475, -0.006614685, -0.001745224, -0.019104004, -0.00982666, 0.025802612, 0.017791748, -0.029403687, -0.0112838745, 0.013931274, -0.021499634, -0.03463745, -0.0032100677, -0.02168274, ...
http://arxiv.org/abs/2301.03225v1
Online Fake Review Detection Using Supervised Machine Learning And BERT Model
Abrar Qadir Mir, Furqan Yaqub Khan, Mohammad Ahsan Chishti
cs.CL, cs.AI, cs.IR
Online shopping stores have grown steadily over the past few years. Due to the massive growth of these businesses, the detection of fake reviews has attracted attention. Fake reviews are seriously trying to mislead customers and thereby undermine the honesty and authenticity of online shopping environments. So far, var...
http://arxiv.org/pdf/2301.03225v1
2023-01-09T09:40:56
2023-01-09T09:40:56
{ "title": [ -0.0002925396, -0.0025863647, -0.0018987656, -0.007106781, -0.004081726, -0.0068855286, -0.0046539307, -0.040405273, -0.051116943, 0.026763916, -0.016845703, 0.0158844, -0.002937317, -0.018463135, -0.01285553, -0.022918701, -0.04043579, ...
http://arxiv.org/abs/2301.03222v1
Machine Learning Algorithms for Depression Detection and Their Comparison
Danish Muzafar, Furqan Yaqub Khan, Mubashir Qayoom
cs.CL, cs.AI, cs.IR
Textual emotional intelligence is playing a ubiquitously important role in leveraging human emotions on social media platforms. Social media platforms are privileged with emotional content and are leveraged for various purposes like opinion mining, emotion mining, and sentiment analysis. This data analysis is also leve...
http://arxiv.org/pdf/2301.03222v1
2023-01-09T09:34:38
2023-01-09T09:34:38
{ "title": [ 0.007881165, -0.0032482147, -0.0041618347, -0.03466797, 0.0067634583, -0.015823364, -0.00050878525, -0.027160645, -0.011779785, 0.03768921, -0.019454956, -0.015899658, 0.01121521, -0.033935547, -0.04486084, -0.0044403076, 0.0007147789, ...
http://arxiv.org/abs/2301.03141v2
Applying Automated Machine Translation to Educational Video Courses
Linden Wang
cs.CL
We studied the capability of automated machine translation in the online video education space by automatically translating Khan Academy videos with state-of-the-art translation models and applying text-to-speech synthesis and audio/video synchronization to build engaging videos in target languages. We also analyzed an...
http://arxiv.org/pdf/2301.03141v2
2023-09-19T01:16:00
2023-01-09T01:44:29
{ "title": [ -0.0262146, -0.0065574646, -0.016998291, -0.01625061, -0.008323669, -0.0058250427, -0.005783081, -0.020950317, -0.012718201, 0.06021118, -0.02659607, -0.0034599304, 0.030319214, -0.013931274, -0.035308838, -0.003124237, -0.008590698, ...
http://arxiv.org/abs/2301.03136v2
Removing Non-Stationary Knowledge From Pre-Trained Language Models for Entity-Level Sentiment Classification in Finance
Guijin Son, Hanwool Lee, Nahyeon Kang, Moonjeong Hahm
cs.CL, cs.LG, q-fin.GN
Extraction of sentiment signals from news text, stock message boards, and business reports, for stock movement prediction, has been a rising field of interest in finance. Building upon past literature, the most recent works attempt to better capture sentiment from sentences with complex syntactic structures by introduc...
http://arxiv.org/pdf/2301.03136v2
2023-01-25T02:06:13
2023-01-09T01:26:55
{ "title": [ -0.004589081, -0.002117157, -0.048919678, -0.030792236, 0.0003387928, -0.026931763, -0.016998291, -0.007888794, 0.023651123, 0.048736572, -0.021148682, -0.025939941, 0.013191223, -0.005290985, -0.08856201, -0.006839752, -0.004928589, ...
http://arxiv.org/abs/2301.03127v2
Logically at Factify 2: A Multi-Modal Fact Checking System Based on Evidence Retrieval techniques and Transformer Encoder Architecture
Pim Jordi Verschuuren, Jie Gao, Adelize van Eeden, Stylianos Oikonomou, Anil Bandhakavi
cs.CL, cs.AI, cs.CV, cs.MM
In this paper, we present the Logically submissions to De-Factify 2 challenge (DE-FACTIFY 2023) on the task 1 of Multi-Modal Fact Checking. We describes our submissions to this challenge including explored evidence retrieval and selection techniques, pre-trained cross-modal and unimodal models, and a cross-modal veraci...
http://arxiv.org/pdf/2301.03127v2
2023-02-03T22:55:12
2023-01-09T00:19:11
{ "title": [ -0.009963989, 0.011726379, -0.04458618, -0.05380249, 0.009132385, -0.005466461, -0.06542969, -0.01701355, -0.022964478, 0.05407715, -0.035736084, 0.030807495, 0.03643799, -0.012794495, -0.022537231, -0.009231567, -0.021697998, 0.0604...
http://arxiv.org/abs/2301.03119v2
Automatic Generation of German Drama Texts Using Fine Tuned GPT-2 Models
Mariam Bangura, Kristina Barabashova, Anna Karnysheva, Sarah Semczuk, Yifan Wang
cs.CL
This study is devoted to the automatic generation of German drama texts. We suggest an approach consisting of two key steps: fine-tuning a GPT-2 model (the outline model) to generate outlines of scenes based on keywords and fine-tuning a second model (the generation model) to generate scenes from the scene outline. The...
http://arxiv.org/pdf/2301.03119v2
2023-01-10T14:08:00
2023-01-08T23:12:46
{ "title": [ -0.039367676, -0.008262634, -0.032684326, -0.024032593, -0.025909424, 0.01423645, -0.04901123, 0.00015199184, -0.009094238, 0.061065674, 0.022415161, 0.000118255615, 0.03704834, -0.021087646, -0.040496826, 0.02420044, -0.022125244, 0...
http://arxiv.org/abs/2301.03095v1
MEGAnno: Exploratory Labeling for NLP in Computational Notebooks
Dan Zhang, Hannah Kim, Rafael Li Chen, Eser Kandogan, Estevam Hruschka
cs.HC, cs.CL
We present MEGAnno, a novel exploratory annotation framework designed for NLP researchers and practitioners. Unlike existing labeling tools that focus on data labeling only, our framework aims to support a broader, iterative ML workflow including data exploration and model development. With MEGAnno's API, users can pro...
http://arxiv.org/pdf/2301.03095v1
2023-01-08T19:16:22
2023-01-08T19:16:22
{ "title": [ -0.045806885, -0.025161743, -0.033813477, -0.041381836, -0.018157959, 0.009346008, -0.051940918, -0.01008606, 0.005092621, 0.017105103, -0.02494812, 0.030685425, -0.0028324127, -0.009941101, -0.034484863, 0.00037240982, -0.028305054, ...
http://arxiv.org/abs/2301.03056v1
The State of Human-centered NLP Technology for Fact-checking
Anubrata Das, Houjiang Liu, Venelin Kovatchev, Matthew Lease
cs.CL, cs.AI, cs.CY, cs.HC
Misinformation threatens modern society by promoting distrust in science, changing narratives in public health, heightening social polarization, and disrupting democratic elections and financial markets, among a myriad of other societal harms. To address this, a growing cadre of professional fact-checkers and journalis...
http://arxiv.org/pdf/2301.03056v1
2023-01-08T15:13:13
2023-01-08T15:13:13
{ "title": [ 0.0041160583, -0.0096206665, -0.035217285, -0.05758667, -0.021087646, -0.0256958, -0.010597229, 0.0009822845, -0.030166626, 0.041870117, -0.050994873, 0.009712219, 0.03845215, -0.007144928, -0.022521973, -0.0014104843, -0.02218628, 0...
http://arxiv.org/abs/2301.03029v6
Topic Modelling of Swedish Newspaper Articles about Coronavirus: a Case Study using Latent Dirichlet Allocation Method
Bernadeta Griciūtė, Lifeng Han, Goran Nenadic
cs.CL, cs.SI
Topic Modelling (TM) is from the research branches of natural language understanding (NLU) and natural language processing (NLP) that is to facilitate insightful analysis from large documents and datasets, such as a summarisation of main topics and the topic changes. This kind of discovery is getting more popular in re...
http://arxiv.org/pdf/2301.03029v6
2023-04-18T16:43:11
2023-01-08T12:33:58
{ "title": [ 0.01689148, 0.0038452148, -0.029373169, -0.04534912, -0.004180908, -0.010505676, -0.0107421875, -0.021194458, -0.025375366, 0.02255249, -0.020568848, 0.068359375, -0.01852417, -0.013221741, 0.008155823, -0.0019187927, -0.0038967133, ...
http://arxiv.org/abs/2301.03012v1
Analyzing the Representational Geometry of Acoustic Word Embeddings
Badr M. Abdullah, Dietrich Klakow
cs.CL
Acoustic word embeddings (AWEs) are vector representations such that different acoustic exemplars of the same word are projected nearby in the embedding space. In addition to their use in speech technology applications such as spoken term discovery and keyword spotting, AWE models have been adopted as models of spoken-...
http://arxiv.org/pdf/2301.03012v1
2023-01-08T10:22:50
2023-01-08T10:22:50
{ "title": [ -0.014549255, -0.011657715, -0.024520874, -0.042755127, -0.009773254, -0.0072898865, -0.0335083, 0.01676941, 0.010864258, 0.047576904, -0.024520874, 0.0093688965, 0.025024414, -0.005680084, -0.0362854, 0.020858765, -0.0041160583, 0.0...
http://arxiv.org/abs/2301.02998v1
InPars-Light: Cost-Effective Unsupervised Training of Efficient Rankers
Leonid Boytsov, Preksha Patel, Vivek Sourabh, Riddhi Nisar, Sayani Kundu, Ramya Ramanathan, Eric Nyberg
cs.IR, cs.AI, cs.CL
We carried out a reproducibility study of InPars recipe for unsupervised training of neural rankers. As a by-product of this study, we developed a simple-yet-effective modification of InPars, which we called InPars-light. Unlike InPars, InPars-light uses only a freely available language model BLOOM and 7x-100x smaller ...
http://arxiv.org/pdf/2301.02998v1
2023-01-08T08:03:46
2023-01-08T08:03:46
{ "title": [ 0.00680542, -0.0033454895, -0.03564453, -0.042541504, 0.019943237, -0.010124207, -0.006137848, -0.038604736, 0.026824951, 0.012046814, -0.014503479, 0.013137817, -0.013496399, -0.032806396, -0.026397705, -0.0018291473, -0.0071792603, ...
http://arxiv.org/abs/2301.02975v2
Traditional Readability Formulas Compared for English
Bruce W. Lee, Jason Hyung-Jong Lee
cs.CL, cs.LG
Traditional English readability formulas, or equations, were largely developed in the 20th century. Nonetheless, many researchers still rely on them for various NLP applications. This phenomenon is presumably due to the convenience and straightforwardness of readability formulas. In this work, we contribute to the NLP ...
http://arxiv.org/pdf/2301.02975v2
2023-01-10T05:54:50
2023-01-08T04:33:43
{ "title": [ -0.032409668, 0.012275696, -0.019683838, 0.013282776, -0.050811768, -0.003993988, -0.026855469, -0.021087646, -0.014038086, 0.03302002, -0.020263672, -0.025787354, 0.029373169, -0.048919678, 0.011856079, -0.0009937286, 0.023971558, 0...
http://arxiv.org/abs/2301.02966v1
SpeeChain: A Speech Toolkit for Large-Scale Machine Speech Chain
Heli Qi, Sashi Novitasari, Andros Tjandra, Sakriani Sakti, Satoshi Nakamura
cs.CL, cs.LG, eess.AS, 68T10, I.2.7
This paper introduces SpeeChain, an open-source Pytorch-based toolkit designed to develop the machine speech chain for large-scale use. This first release focuses on the TTS-to-ASR chain, a core component of the machine speech chain, that refers to the TTS data augmentation by unspoken text for ASR. To build an efficie...
http://arxiv.org/pdf/2301.02966v1
2023-01-08T03:16:56
2023-01-08T03:16:56
{ "title": [ -0.044921875, -0.0060424805, -0.08331299, -0.08483887, 0.020126343, 0.003030777, -0.06774902, 0.04260254, -0.0063171387, 0.035095215, -0.028274536, -0.007785797, 0.010932922, -0.019943237, -0.037139893, -0.0022964478, -0.008422852, 0...
http://arxiv.org/abs/2301.06901v1
Graph-based Keyword Planning for Legal Clause Generation from Topics
Sagar Joshi, Sumanth Balaji, Aparna Garimella, Vasudeva Varma
cs.CL, cs.AI, I.2.7
Generating domain-specific content such as legal clauses based on minimal user-provided information can be of significant benefit in automating legal contract generation. In this paper, we propose a controllable graph-based mechanism that can generate legal clauses using only the topic or type of the legal clauses. Our...
http://arxiv.org/pdf/2301.06901v1
2023-01-07T15:18:45
2023-01-07T15:18:45
{ "title": [ -0.05731201, -0.019943237, -0.027908325, -0.03479004, -0.035125732, 0.030075073, -0.052124023, 0.002752304, -0.0046195984, 0.02053833, -0.02897644, 0.058135986, 0.002916336, 0.00086832047, -0.0065193176, 0.0151901245, -0.022857666, 0...
http://arxiv.org/abs/2301.02828v2
Why do Nearest Neighbor Language Models Work?
Frank F. Xu, Uri Alon, Graham Neubig
cs.CL, cs.LG
Language models (LMs) compute the probability of a text by sequentially computing a representation of an already-seen context and using this representation to predict the next word. Currently, most LMs calculate these representations through a neural network consuming the immediate previous context. However recently, r...
http://arxiv.org/pdf/2301.02828v2
2023-01-17T08:02:58
2023-01-07T11:12:36
{ "title": [ -0.046295166, -0.014457703, -0.0002539158, 0.0012350082, -0.07019043, 0.0027656555, -0.08087158, -0.02746582, -0.025222778, 0.05493164, -0.022842407, -0.02041626, 0.0034999847, -0.015838623, -0.014671326, -0.020874023, -0.009880066, ...
http://arxiv.org/abs/2301.02809v1
A Brain-inspired Memory Transformation based Differentiable Neural Computer for Reasoning-based Question Answering
Yao Liang, Hongjian Fang, Yi Zeng, Feifei Zhao
cs.AI, cs.CL
Reasoning and question answering as a basic cognitive function for humans, is nevertheless a great challenge for current artificial intelligence. Although the Differentiable Neural Computer (DNC) model could solve such problems to a certain extent, the development is still limited by its high algorithm complexity, slow...
http://arxiv.org/pdf/2301.02809v1
2023-01-07T08:39:57
2023-01-07T08:39:57
{ "title": [ -0.023391724, -0.00017988682, -0.02557373, -0.035888672, -0.0059890747, -0.0016803741, -0.008735657, 0.007007599, 0.0060920715, 0.03265381, -0.024017334, -0.021057129, 0.017837524, -0.022216797, -0.04208374, -0.011947632, 0.0028476715, ...