Article
Computer Science, Artificial Intelligence
Wisha Zehra, Abdul Rehman Javed, Zunera Jalil, Habib Ullah Khan, Thippa Reddy Gadekallu
Summary: The paper discusses how to achieve cross-corpus, multi-lingual speech emotion recognition system using a majority voting technique to replace traditional single classifier approach. Experimental validation shows that different classifiers give the highest accuracy for different corpora, while the ensemble learning approach combines the effects of various classifiers rather than choosing a single classifier.
COMPLEX & INTELLIGENT SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Matus Pikuliak, Marian Simko, Maria Bielikova
Summary: This study surveys 173 cross-lingual learning papers in text processing, examining tasks, data sets, and languages used. The key contribution is the identification and analysis of four types of cross-lingual transfer based on what is being transferred, which can assist NLP researchers in understanding how to apply cross-lingual learning to various problems. Additionally, important research directions are highlighted to guide future work in cross-lingual learning, aiming to provide a comprehensive overview of the field.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Riikka Huusari, Cecile Capponi, Paul Villoutreix, Hachem Kadri
Summary: This article discusses the problem of kernel completion in the presence of multiple views in the data. The proposed Cross-View Kernel Transfer (CVKT) procedure completes the kernel matrices by aligning the features of other views to represent the target view. The missing values in the kernel matrices can be predicted using data from other views. Simulated and real datasets demonstrate the benefits of the approach.
PATTERN RECOGNITION
(2022)
Article
Computer Science, Information Systems
Juuso Eronen, Michal Ptaszynski, Fumito Masui, Masaki Arata, Gniewosz Leliwa, Michal Wroczynski
Summary: This study demonstrates the effectiveness of cross-lingual transfer learning for zero-shot abusive language detection, showing that linguistic similarity is correlated with classifier performance. The research allows for the use of existing data from higher-resource languages to improve detection systems for low-resource languages.
INFORMATION PROCESSING & MANAGEMENT
(2022)
Article
Computer Science, Artificial Intelligence
Andraz Pelicon, Ravi Shekhar, Blaz Skrlj, Matthew Purver, Senja Pollak
Summary: Datasets, shared tasks, and models have been proposed for various languages but good accuracy in automatic detection relies on substantial, well-labelled datasets. Transfer learning has the potential to reduce the need for task-specific labeled data, but most research focuses on English, posing challenges for languages with few labeled datasets.
PEERJ COMPUTER SCIENCE
(2021)
Article
Computer Science, Information Systems
Juuso Eronen, Michal Ptaszynski, Fumito Masui
Summary: This study focuses on the selection of transfer languages for different NLP tasks. It proposes using linguistic similarity metrics to measure language distance and choose the optimal transfer language. The study demonstrates that linguistic similarity is correlated with cross-lingual transfer performance and that there is a statistically significant difference in choosing the optimal transfer source language. The results show the potential for leveraging knowledge from high-resource languages to improve language applications with limited data.
INFORMATION PROCESSING & MANAGEMENT
(2023)
Article
Computer Science, Artificial Intelligence
Jiqian Mo, Zhiguo Gong
Summary: This article proposes a novel Cross-city Multi-Granular Adaptive Transfer Learning method (MGAT) for traffic prediction. By training the model on multiple source cities and obtaining multi-granular features, the Adaptive Transfer module selects the most appropriate features to improve traffic prediction.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Automation & Control Systems
Guanjin Wang, Kup-Sze Choi, Jeremy Yuen-Chun Teoh, Jie Lu
Summary: This article introduces a new approach called DCOT-LS-SVMs, which is based on least-squares support vector machines and utilizes deep cross-output knowledge transfer. The approach improves the generalizability of LS-SVMs and simplifies the parameter tuning process. Experimental results on UCI datasets and a prostate cancer diagnosis case study demonstrate the effectiveness of the proposed approach.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Article
Computer Science, Artificial Intelligence
Ander Barrena, Aitor Soroa, Eneko Agirre
Summary: This study introduces a zero-shot XNED architecture, eliminating the need for native prior probabilities by having a model for each possible mention string, resulting in significant improvements on XNED datasets in Spanish and Chinese.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Rosario Catelli, Luca Bevilacqua, Nicola Mariniello, Vladimiro Scotto di Carlo, Massimo Magaldi, Hamido Fujita, Giuseppe De Pietro, Massimo Esposito
Summary: The study demonstrates that using transfer learning techniques from languages with high resources to languages with low resources can significantly improve performance, and a multilingual BERT model fine-tuned on English/Italian dataset outperforms language-specific models.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Alejandro Moreo, Andrea Pedrotti, Fabrizio Sebastiani
Summary: Funnelling is a method for cross-lingual text classification based on a two-tier learning ensemble. It uses a meta-classifier to exploit class-class correlations, giving it an edge over other systems.
ACM TRANSACTIONS ON INFORMATION SYSTEMS
(2023)
Article
Computer Science, Information Systems
Rizka W. Sholikah, Agus Z. Arifin, Chastine Fatichah, Ayu Purwarianti
Summary: Semantic relation detection is crucial in natural language processing. This paper proposes a framework for cross-lingual semantic relation identification using multi-task learning and a general vector space, which overcomes the limitation of cross-lingual data and improves relation identification accuracy.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Chemistry, Multidisciplinary
Chanhee Lee, Kisu Yang, Taesun Whang, Chanjun Park, Andrew Matteson, Heuiseok Lim
Summary: This study improves data efficiency by pretraining language models on high-resource languages and treating language modeling of low-resource languages as a domain adaptation task. By selectively reusing parameters from high-resource language models and post-training them alongside learning language-specific parameters in low-resource languages, the method outperforms monolingual training in intrinsic and extrinsic evaluations.
APPLIED SCIENCES-BASEL
(2021)
Article
Automation & Control Systems
Deqing Wang, Junjie Wu, Jingyuan Yang, Baoyu Jing, Wenjie Zhang, Xiaonan He, Hui Zhang
Summary: This article introduces a semisupervised transfer learning approach called ssSCL-ST, which uses structural correspondence learning and space transfer to improve cross-lingual sentiment analysis, aiming to explore intrinsic sentiment knowledge in the target language domain and enhance classification accuracy.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Article
Computer Science, Information Systems
Hanqian Wu, Zhike Wang, Feng Qing, Shoushan Li
Summary: This paper proposes a method for conducting Cross-Lingual Aspect Sentiment Classification (CLASC) task by leveraging rich resources in a source language for aspect sentiment classification in a target language, using bilingual lexicon and machine translation tools for sentence translation. The proposed approach combines Reinforced Transformer with Cross-Lingual Distillation to mitigate the undesirable effects of translation ambiguities in different target languages, outperforming existing methods in experimental results.
Article
Computer Science, Artificial Intelligence
Lin Gui, Yulan He
Summary: This research introduces a joint learning framework for simultaneous unsupervised aspect extraction at the sentence level and supervised sentiment classification at the document level. The framework achieved high sentiment classification accuracy when tested on healthcare services reviews, outperforming several strong baselines.
ARTIFICIAL INTELLIGENCE IN MEDICINE
(2021)
Article
Computer Science, Artificial Intelligence
Bin Liang, Hang Su, Lin Gui, Erik Cambria, Ruifeng Xu
Summary: This paper proposes a graph convolutional network model Sentic GCN based on SenticNet to enhance the affective dependencies of sentences for aspect-based sentiment analysis. By integrating emotional knowledge from SenticNet, the model effectively handles contextual affective information in sentences, improving the effectiveness of sentiment polarity detection towards specific aspects.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Information Systems
Bin Liang, Xiang Li, Lin Gui, Yonghao Fu, Yulan He, Min Yang, Ruifeng Xu
Summary: Existing sentiment analysis methods in aspect-based/category focus successfully detect sentiment polarity towards fixed aspect categories. However, practical applications involve changing aspect categories. Dealing with unseen categories is not fully explored in current methods. In this article, we propose a few-shot aspect category sentiment analysis task and introduce a novel Aspect-Focused Meta-Learning (AFML) framework to effectively predict sentiment polarity of unseen aspect categories.
ACM TRANSACTIONS ON INFORMATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Lin Gui, Leng Jia, Jiyun Zhou, Ruifeng Xu, Yulan He
Summary: This paper proposes a multi-task learning framework that jointly learns a sentiment classifier and a topic model, aiming to make the word-level latent topic distributions in the topic model similar to the word-level attention vectors in sentiment classifiers. The experimental results on Yelp and IMDB datasets demonstrate the superior performance of the proposed framework in both sentiment classification and topic modeling tasks.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2022)
Article
Computer Science, Information Systems
Zhiyuan Wen, Lin Gui, Qianlong Wang, Mingyue Guo, Xiaoqi Yu, Jiachen Du, Ruifeng Xu
Summary: Sarcasm expression is a literary technique where people intentionally express the opposite of what is implied. Accurate detection of sarcasm can help understand speakers' true intentions and improve other natural language processing tasks. Detecting sarcasm in Chinese is more challenging due to the characteristics of the language. To address this, a sememe and auxiliary enhanced attention neural model is proposed.
INFORMATION PROCESSING & MANAGEMENT
(2022)
Article
Computer Science, Artificial Intelligence
Hanqi Yan, Lin Gui, Yulan He
Summary: In recent years, there has been increasing interest in developing interpretable models in Natural Language Processing (NLP). However, it is difficult to accurately explain model decisions by words or phrases when neural models in NLP compose word semantics hierarchically. This article proposes a novel Hierarchical Interpretable Neural Text classifier, called HINT, which generates explanations of model predictions in the form of label-associated topics. Experimental results show that HINT achieves comparable text classification results and provides better interpretations than other interpretable neural text classifiers.
COMPUTATIONAL LINGUISTICS
(2022)
Article
Computer Science, Artificial Intelligence
Bin Liang, Rongdi Yin, Jiachen Du, Lin Gui, Yulan He, Min Yang, Ruifeng Xu
Summary: The state-of-the-art approaches to targeted aspect-based sentiment analysis (TABSA) are mostly built on deep neural networks with attention mechanisms. One problem is that embeddings of targets and aspects are either pre-trained from large external corpora or randomly initialized. We propose an embedding refinement framework called RAEC (Refining Affective Embedding from Context) that incorporates affective commonsense knowledge and word relative location information to derive context-affective embeddings. Experimental results show that our framework achieves state-of-the-art results compared to models relying on pre-trained word embeddings or built on other embedding refinement methods.
IEEE TRANSACTIONS ON AFFECTIVE COMPUTING
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Bin Liang, Qinglin Zhu, Xiang Li, Min Yang, Lin Gui, Yulan He, Ruifeng Xu
Summary: This paper proposes a joint contrastive learning framework for zero-shot stance detection, which achieves state-of-the-art performance by generalizing stance features and reasoning ability for unseen targets through stance contrastive learning and target-aware prototypical graph contrastive learning.
PROCEEDINGS OF THE 60TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2022), VOL 1: (LONG PAPERS)
(2022)
Proceedings Paper
Computer Science, Cybernetics
Bin Liang, Zixiao Chen, Lin Gui, Yulan He, Min Yang, Ruifeng Xu
Summary: This paper proposes a framework for zero-shot stance detection that effectively distinguishes the types of stance features and learns transferable features. By treating stance feature type identification as a pretext task and using a hierarchical contrastive learning strategy to capture correlations and differences, the model is able to better represent the stance of previously unseen targets.
PROCEEDINGS OF THE ACM WEB CONFERENCE 2022 (WWW'22)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Bin Liang, Chenwei Lou, Xiang Li, Min Yang, Lin Gui, Yulan He, Wenjie Pei, Ruifeng Xu
Summary: In this paper, the authors investigate multimodal sarcasm detection from a novel perspective by constructing a cross-modal graph to explicitly capture the ironic relations between textual and visual modalities. They propose a cross-modal graph convolutional network which achieves state-of-the-art performance in multimodal sarcasm detection.
PROCEEDINGS OF THE 60TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2022), VOL 1: (LONG PAPERS)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Bin Liang, Yonghao Fu, Lin Gui, Min Yang, Jiachen Du, Yulan He, Ruifeng Xu
Summary: Target plays a crucial role in stance detection, and handling unknown targets is a significant challenge. This paper introduces a novel approach to achieve stance detection across different targets, by constructing target-adaptive dependency graphs specific to each target.
PROCEEDINGS OF THE WORLD WIDE WEB CONFERENCE 2021 (WWW 2021)
(2021)
Proceedings Paper
Computer Science, Information Systems
Chenwei Lou, Bin Liang, Lin Gui, Yulan He, Yixue Dang, Ruifeng Xu
Summary: This paper introduces a novel approach for detecting sarcastic expressions by constructing affective and dependency graphs based on external affective commonsense knowledge and the syntactical information of sentences, proposing an ADGCN framework. Experimental results demonstrate that this method outperforms the current state-of-the-art methods in sarcasm detection.
SIGIR '21 - PROCEEDINGS OF THE 44TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Hanqi Yan, Lin Gui, Gabriele Pergola, Yulan He
Summary: This study focuses on the dataset bias in Emotion Cause Extraction (ECE) task and proposes a new strategy to reduce the dependency on relative positions of clauses by generating adversarial examples. By introducing a graph-based method to explicitly model emotion triggering paths, the model enhances the understanding of semantic dependencies, making it more robust against adversarial attacks compared to existing models.
59TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS AND THE 11TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING (ACL-IJCNLP 2021), VOL 1
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Lixing Zhu, Gabriele Pergola, Lin Gui, Deyu Zhou, Yulan He
Summary: The paper introduces a Topic-Driven Knowledge-Aware Transformer model for emotion detection in dialogues, which combines topic-augmented language model with commonsense knowledge to achieve superior performance in distinguishing emotion categories.
59TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS AND THE 11TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING, VOL 1 (ACL-IJCNLP 2021)
(2021)
Article
Acoustics
Chuang Fan, Chaofa Yuan, Lin Gui, Yue Zhang, Ruifeng Xu
Summary: The study introduces a multi-task sequence tagging framework for extracting emotions and associated causes simultaneously, by encoding distances into a tagging scheme to achieve explicit and interpretable information exchange for emotion-cause pair extraction.
IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING
(2021)