nach oben

Journal of Intelligent Information Systems

Erschienen in:

20.08.2022

Multi-task learning for toxic comment classification and rationale extraction

verfasst von: Kiran Babu Nelatoori, Hima Bindu Kommanti

Erschienen in: Journal of Intelligent Information Systems | Ausgabe 2/2023

Einloggen

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config

KI-gestützte Suche

Aus

Abstract

Social media content moderation is the standard practice as on today to promote healthy discussion forums. Toxic span prediction is helpful for explaining the toxic comment classification labels, thus is an important step towards building automated moderation systems. The relation between toxic comment classification and toxic span prediction makes joint learning objective meaningful. We propose a multi-task learning model using ToxicXLMR for bidirectional contextual embeddings of input text for toxic comment classification, and a Bi-LSTM CRF layer for toxic span or rationale identification. To enable multi-task learning in this domain, we have curated a dataset from Jigsaw and Toxic span prediction datasets. The proposed model outperformed the single task models on the curated and toxic span prediction datasets with 4% and 2% improvement for classification and rationale identification, respectively. We investigated the domain adaptation ability of the proposed MTL model on HASOC and OLID datasets that contain the out of domain text from Twitter and found a 3% improvement in the F1 score over single task models.

Vorheriger Artikel Semantic event relationships identification and representation using HyperGraph in multimedia digital ecosystem

Nächster Artikel A context-aware unsupervised predictive maintenance solution for fleet management

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

über 102.000 Bücher
über 537 Zeitschriften

aus folgenden Fachgebieten:

Automobil + Motoren
Bauwesen + Immobilien
Business IT + Informatik
Elektrotechnik + Elektronik
Energie + Nachhaltigkeit
Finance + Banking
Management + Führung
Marketing + Vertrieb
Maschinenbau + Werkstoffe
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Springer Professional "Technik"

Online-Abonnement

Mit Springer Professional "Technik" erhalten Sie Zugriff auf:

über 67.000 Bücher
über 390 Zeitschriften

aus folgenden Fachgebieten:

Automobil + Motoren
Bauwesen + Immobilien
Business IT + Informatik
Elektrotechnik + Elektronik
Energie + Nachhaltigkeit
Maschinenbau + Werkstoffe

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

über 67.000 Bücher
über 340 Zeitschriften

aus folgenden Fachgebieten:

Bauwesen + Immobilien
Business IT + Informatik
Finance + Banking
Management + Führung
Marketing + Vertrieb
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

https://enough.org/stats_cyberbullying

https://sites.google.com/view/trac2/home

https://social-threats.github.io/

https://www.workshopononlineabuse.com/

https://www.kaggle.com/competitions/jigsaw-unintended-bias-in-toxicity-classification/data

https://hasocfire.github.io/hasoc/2019/dataset.html

https://scholar.harvard.edu/malmasi/olid

https://huggingface.co/unitary/toxic-bert

https://huggingface.co/unitary/unbiased-toxic-roberta

https://huggingface.co/unitary/multilingual-toxic-xlm-roberta

https://www.kaggle.com/competitions/jigsaw-unintended-bias-in-toxicity-classification/data

https://medium.com/@aja_15265/saying-goodbye-to-civil-comments-41859d3a2b1d

https://huggingface.co/

Adadi, A., & Berrada, M. (2018). Peeking inside the black-box: A survey on explainable artificial intelligence (xai). IEEE Access, 6, 52,138–52,160. https://doi.org/10.1109/ACCESS.2018.2870052CrossRef

Ahmed, SS, & Kumar M., A. (2021). Classification of censored tweets in Chinese language using XLNet. In Proceedings of the fourth workshop on NLP for internet freedom: Censorship, disinformation, and propaganda. https://doi.org/10.18653/v1/2021.nlp4if-1.21 (pp. 136–139).

Akhtar, M.S., Garg, T., & Ekbal, A. (2020). Multi-task learning for aspect term extraction and aspect sentiment classification. Neurocomputing, 398, 247–256. https://doi.org/10.1016/j.neucom.2020.02.093CrossRef

Ashok Kumar, J, Abirami, S, Tina Esther, T., & et al (2021). Comment toxicity detection via a multichannel convolutional bidirectional gated recurrent unit. Neurocomputing, 441, 272–278. https://doi.org/10.1016/j.neucom.2021.02.023CrossRef

Badjatiya, P., Gupta, S., Gupta, M., & et al (2017). Deep learning for hate speech detection in tweets. In Proceedings of the 26th international conference on world wide web companion. https://doi.org/10.1145/3041021.3054223 (pp. 759–760).

Bansal, A., Kaushik, A., & Modi, A. (2021). IITK@detox at SemEval-2021 task 5: Semi-supervised learning and dice loss for toxic spans detection. In Proceedings of the 15th international workshop on semantic evaluation (SemEval-2021). https://doi.org/10.18653/v1/2021.semeval-1.24 (pp. 211–219).

Baxter, J. (2000). A model of inductive bias learning. Journal of artificial intelligence research, 12, 149–198. https://doi.org/10.5555/1622248.1622254CrossRefMATHMathSciNet

Caruana, R. (1997). Multitask learning. Machine learning, 28(1), 41–75. https://doi.org/10.1023/A:1007379606734CrossRefMathSciNet

Caruana, R.A (1993). Multitask connectionist learning. In Proceedings of the 1993 connectionist models summer school.

Caselli, T., Basile, V., Mitrović, J., & et al (2021). HateBERT: Retraining BERT for abusive language detection in English. In Proceedings of the 5th workshop on online abuse and harms (WOAH 2021). https://doi.org/10.18653/v1/2021.woah-1.3(pp. 17–25).

Chakrabarty, T., Gupta, K., & Muresan, S. (2019). Pay “attention” to your context when classifying abusive language. In Proceedings of the third workshop on abusive language online. https://doi.org/10.18653/v1/W19-3508 (pp. 70–79).

Chen, Q., Zhuo, Z., & Wang, W. (2019). Bert for joint intent classification and slot filling. arXiv:1902.10909

Chia, Z.L., Ptaszynski, M., Masui, F., & et al (2021). Machine learning and feature engineering-based study into sarcasm and irony classification with application to cyberbullying detection. Information Processing and Management, 58, 102600. https://doi.org/10.1016/j.ipm.2021.102600CrossRef

Conneau, A., Khandelwal, K., Goyal, N., & et al (2020). Unsupervised cross-lingual representation learning at scale. In Proceedings of the 58th annual meeting of the association for computational linguistics. https://doi.org/10.18653/v1/2020.acl-main.747 (pp. 8440–8451).

Da San Martino, G., Yu, S., Barrón-Cedeño, A., & et al (2019). Fine-grained analysis of propaganda in news article. In Proceedings of the 2019 conference on empirical methods in natural language processing and the 9th international joint conference on natural language processing (EMNLP-IJCNLP). https://doi.org/10.18653/v1/D19-1565 (pp. 5636–5646).

Davidson, T., Warmsley, D., Macy, M., & et al (2017). Automated hate speech detection and the problem of offensive language. In Proceedings of the international AAAI conference on web and social media. https://ojs.aaai.org/index.php/ICWSM/article/view/14955 (pp. 512–515).

Dellerman, D. (2022). Influence of cyberbullying on suicidal behaviors. Ph.D. Thesis, Walden University.

Devlin, J., Chang, M.-W., Lee, K., & et al (2019). BERT: pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 conference of the North American chapter of the association for computational linguistics: human language technologies, volume 1 (Long and Short Papers). https://doi.org/10.18653/v1/N19-1423 (pp. 4171–4186).

Ed-drissiya, E, Sarrouti, M., En-Nahnahi, N., & et al (2021). Mttlade: A multi-task transfer learning-based method for adverse drug events extraction. Information Processing and Management, 58(3), 102473. https://doi.org/10.1016/j.ipm.2020.102473CrossRef

Elnaggar, A., Waltl, B., Glaser, I., & et al (2018). Stop illegal comments: A multi-task deep learning approach. In Proceedings of the 2018 artificial intelligence and cloud computing conference. https://doi.org/10.1145/3299819.3299845 (pp. 41–47).

Fortuna, P., Soler-Company, J., & Wanner, L. (2021). How well do hate speech, toxicity, abusive and offensive language classification models generalize across datasets? Information Processing and Management, 58(3), 102524. https://doi.org/10.1016/j.ipm.2021.102524CrossRef

Founta, A.M., Chatzakou, D., Kourtellis, N., & et al (2019). A unified deep learning architecture for abuse detection. In Proceedings of the 10th ACM conference on web science. https://doi.org/10.1145/3292522.3326028 (pp. 105–114).

Gong, T., Lee, T., Stephenson, C., & et al (2019). A comparison of loss weighting strategies for multi task learning in deep neural networks. IEEE Access, 7, 141627–141632. https://doi.org/10.1109/ACCESS.2019.2943604CrossRef

Huang, Z., Xu, W., & Yu, K. (2015). Bidirectional lstm-crf models for sequence tagging. arXiv:1508.01991

Karen, S., Andrea, V., & Andrew, Z. (2014). Deep inside convolutional networks: visualising image classification models and saliency maps. arXiv:1312.6034

Khan, Y., Ma, W., & Vosoughi, S. (2021). Lone pine at SemEval-2021 task 5: fine-grained detection of hate speech using BERToxic. In Proceedings of the 15th international workshop on semantic evaluation (SemEval-2021). https://doi.org/10.18653/v1/2021.semeval-1.132 (pp. 967–973).

Kiran Babu, N., & HimaBindu, K. (2022). Attention-based bi-lstm network for abusive language detection. IETE Journal of Research, 1–9. https://doi.org/10.1080/03772063.2022.2034534

Liu, P., Li, W., & Zou, L. (2019a). NULI at SemEval-2019 task 6: Transfer learning for offensive language detection using bidirectional transformers. In Proceedings of the 13th international workshop on semantic evaluation. https://doi.org/10.18653/v1/S19-2011 (pp. 87–91).

Liu, Y., Ott, M., Goyal, N., & et al. (2019b). Roberta: A robustly optimized bert pretraining approach. arXiv:1907.11692

Long, M., Cao, Z., Wang, J., & et al (2017). Learning multiple tasks with multilinear relationship networks. In Proceedings of the 31st international conference on neural information processing systems. https://dl.acm.org/doi/10.5555/3294771.3294923(pp. 1593–1602).

Ma, X., & Hovy, E. (2016). End-to-end sequence labeling via bi-directional LSTM-CNNs-CRF. In Proceedings of the 54th annual meeting of the association for computational linguistics (volume 1: long papers). https://doi.org/10.18653/v1/P16-1101(pp. 1064–1074).

Mathew, B., Saha, P., Yimam, SM, & et al (2021). Hatexplain: a benchmark dataset for explainable hate speech detection, 14,867–14,875. https://ojs.aaai.org/index.php/AAAI/article/view/17745

McCann, B., Keskar, N.S., Xiong, C., & et al. (2018). The natural language decathlon: Multitask learning as question answering. arXiv:1806.08730

Mehta, D., Dwivedi, A., Patra, A., & et al (2021). A transformer-based architecture for fake news classification. Social Network Analysis and Mining, 11(1), 1–12. https://doi.org/10.1007/s13278-021-00738-yCrossRef

Mozafari, M., Farahbakhsh, R., & Crespi, N. (2019). A BERT-based transfer learning approach for hate speech detection in online social media. In Complex networks 2019: 8th international conference on complex networks and their applications. https://doi.org/10.1007/978-3-030-36687-277 (pp. 928–940).

Nachar, N. (2008). The mann-whitney u: A test for assessing whether two independent samples come from the same distribution. Tutorials in Quantitative Methods for Psychology, 4, 13–20. https://doi.org/10.20982/tqmp.04.1.p013CrossRef

Nguyen, VA, Nguyen, TM, Quang Dao, H., & et al (2021). S-NLP at SemEval-2021 task 5: An analysis of dual networks for sequence tagging. In Proceedings of the 15th international workshop on semantic evaluation (SemEval-2021). https://doi.org/10.18653/v1/2021.semeval-1.120 (pp. 888–897).

Pamungkas, EW, & Patti, V. (2019). Cross-domain and cross-lingual abusive language detection: A hybrid approach with deep learning and a multilingual lexicon. In Proceedings of the 57th annual meeting of the association for computational linguistics: student research workshop. https://doi.org/10.18653/v1/P19-2051 (pp. 363–370).

Park, JH, & Fung, P. (2017). One-step and two-step classification for abusive language detection on Twitter. In Proceedings of the first workshop on abusive language online. https://doi.org/10.18653/v1/W17-3006 (pp. 41–45). Vancouver: Association for Computational Linguistics.

Pavlopoulos, J., Sorensen, J., Laugier, L, & et al (2021). SemEval-2021 task 5: Toxic spans detection. In Proceedings of the 15th international workshop on semantic evaluation (SemEval-2021). https://doi.org/10.18653/v1/2021.semeval-1.6 (pp. 59–69).

Ramsundar, B., Kearnes, S., Riley, P., & et al. (2015). Massively multitask networks for drug discovery. arXiv:1502.02072

Ribeiro, MT, Singh, S., & Guestrin, C. (2016). “Why should i trust you?” Explaining the predictions of any classifier. In Proceedings of the ACM SIGKDD international conference on knowledge discovery and data mining. https://doi.org/10.1145/2939672.2939778 (pp. 1135–1144).

Ruder, S. (2017). An overview of multi-task learning in deep neural networks. arXiv:1706.05098

Schuster, M., & Nakajima, K. (2012). Japanese and korean voice search. In 2012 IEEE International conference on acoustics, speech and signal processing (ICASSP). https://doi.org/10.1109/ICASSP.2012.6289079 (pp. 5149–5152).

Sharma, M., Kandasamy, I., & Vasantha, W.b. (2021). YoungSheldon at SemEval-2021 task 5: Fine-tuning pre-trained language models for toxic spans detection using token classification objective. In Proceedings of the 15th international workshop on semantic evaluation (SemEval-2021). https://doi.org/10.18653/v1/2021.semeval-1.130(pp. 953–959).

Sonone, S. S, Sankhla, MS, & Kumar, R. (2021). Cyber bullying. In Combating the exploitation of children in cyberspace: emerging research and opportunities. https://doi.org/10.4018/978-1-7998-2360-5.ch001 (pp. 1–18).

Standley, T., Zamir, A., Chen, D., & et al (2020). Which tasks should be learned together in multi-task learning?. In Proceedings of the 37th international conference on machine learning. https://proceedings.mlr.press/v119/standley20a.html(pp. 9120–9132).

Sundararajan, M., Taly, A., & Yan, Q. (2017). Axiomatic attribution for deep networks. In Proceedings of the 34th international conference on machine learning. https://dl.acm.org/doi/10.5555/3305890.3306024 (pp. 3319–3328).

Temper, M., Poisel, R., & Tjoa, S. (2013). Facebook watchdog: A research agenda for detecting online grooming and bullying activities. In IEEE International conference on systems, man, and cybernetics, SMC. https://doi.org/10.1109/SMC.2013.487 (pp. 2854–2859).

Van Aken, B., Risch, J., Krestel, R., & et al (2018). Challenges for toxic comment classification: An in-depth error analysis. In Proceedings of the 2nd workshop on abusive language online (ALW2). https://doi.org/10.18653/v1/W18-5105 (pp. 33–42).

Vaswani, A., Shazeer, N., Parmar, N., & et al (2017). Attention is all you need. In Advances in neural information processing systems. https://dl.acm.org/doi/10.5555/3295222.3295349

Viterbi, A. J. (2009). Viterbi algorithm. Scholarpedia, 4(1), 6246. https://doi.org/10.4249/scholarpedia.6246CrossRef

Wang, B., Ding, Y., Liu, S., & Zhou, X. (2019). Ynu_wb at HASOC 2019: Ordered neurons LSTM with attention for identifying hate speech and offensive language. In Working notes of FIRE 2019 - forum for information retrieval evaluation. http://ceur-ws.org/Vol-2517/T3-2.pdf (pp. 191–198).

Wang, X., Xu, G., Zhang, Z., & et al (2021). End-to-end aspect-based sentiment analysis with hierarchical multi-task learning. Neurocomputing, 455, 178–188. https://doi.org/10.1016/j.neucom.2021.03.100CrossRef

Waseem, Z., & Hovy, D. (2016). Hateful symbols or hateful people? Predictive features for hate speech detection on Twitter. In Proceedings of the NAACL student research workshop. https://doi.org/10.18653/v1/N16-2013 (pp. 88–93).

Wiegreffe, S., & Pinter, Y. (2019). Attention is not not explanation. In Proceedings of the 2019 conference on empirical methods in natural language processing and the 9th international joint conference on natural language processing (EMNLP-IJCNLP). https://doi.org/10.18653/v1/D19-1002 (pp. 11–20).

Worsham, J., & Kalita, J. (2020). Multi-task learning for natural language processing in the 2020s: where are we going? Pattern Recognition Letters, 136, 120–126. https://doi.org/10.1016/j.patrec.2020.05.031CrossRef

Xiang, T., Macavaney, S., Yang, E., & et al (2021). Toxccin: Toxic content classification with interpretability. In Proceedings of the 11th workshop on computational approaches to subjectivity, sentiment and social media analysis. https://aclanthology.org/2021.wassa-1.1 (pp. 1–12).

Xu, K., Ba, J., Kiros, R., & et al (2015). Show, attend and tell: Neural image caption generation with visual attention. In Proceedings of the 32nd international conference on machine learning. https://doi.org/10.5555/3045118.3045336(pp. 2048–2057).

Zaidan, O., Eisner, J., & Piatko, C. (2007). Using “annotator rationales” to improve machine learning for text categorization. In Human language technologies 2007: the conference of the North American chapter of the association for computational linguistics; proceedings of the main conference. https://aclanthology.org/N07-1033 (pp. 260–267).

Zhang, Z., Robinson, D., & Tepper, J. (2018). Detecting hate speech on twitter using a convolution-gru based deep neural network. In European semantic web conference. https://doi.org/10.1007/978-3-319-93417-4∖_48 (pp. 745–760).

Zhu, Q., Lin, Z., Zhang, Y., & et al (2021). HITSZ-HLT at SemEval-2021 task 5: Ensemble sequence labeling and span boundary detection for toxic span detection. In Proceedings of the 15th international workshop on semantic evaluation (SemEval-2021). https://doi.org/10.18653/v1/2021.semeval-1.63 (pp. 521–526).

Zou, L., & Li, W. (2021). LZ1904 at SemEval-2021 task 5: Bi-LSTM-CRF for toxic span detection using pretrained word embedding. In Proceedings of the 15th international workshop on semantic evaluation (SemEval-2021). https://doi.org/10.18653/v1/2021.semeval-1.138 (pp. 1009–1014).

Titel: Multi-task learning for toxic comment classification and rationale extraction
verfasst von: Kiran Babu Nelatoori
Hima Bindu Kommanti
Publikationsdatum: 20.08.2022
Verlag: Springer US
Erschienen in: Journal of Intelligent Information Systems / Ausgabe 2/2023
Print ISSN: 0925-9902
Elektronische ISSN: 1573-7675
DOI: https://doi.org/10.1007/s10844-022-00726-4

Springer Professional

Abstract

Bitte loggen Sie sich ein, um Zugang zu Ihrer Lizenz zu erhalten.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Springer Professional "Technik"

Springer Professional "Wirtschaft"

Weitere Artikel der Ausgabe 2/2023

Hierarchical reinforcement learning for efficient and effective automated penetration testing of large networks

A context-aware unsupervised predictive maintenance solution for fleet management

Machine learning in industrial control system (ICS) security: current landscape, opportunities and challenges

DoWTS – Denial-of-Wallet Test Simulator: Synthetic data generation for preemptive defence

Windows and IoT malware visualization and classification with deep CNN and Xception CNN using Markov images

Leveraging siamese networks for one-shot intrusion detection model

Premium Partner