A Contextualized Embeddings-based Method to Detect Suicide Ideations in Texts

• Wallace Da Silva Coelho Valadão IME
• Eric Rodrigues Da Silva IME
• Paulo Márcio Souza Freire IME
• Ronaldo Ribeiro Goldschmidt IME

Resumo

Nowadays, suicide is one of the leading causes of death for young people worldwide. Many of those youngsters expose their suicidal intentions on social media. Prevention based on suicide ideation (SI) detection in social media posts is an important strategy to avoid the occurrence of this type of death. Although several studies have developed methods to automatically detect SI in texts, as far as it was possible to observe, none of them uses contextualized embeddings (i.e. vector representations of texts that consider the context where words and sentences occur). Therefore, the present work hypothesizes that representing texts with contextualized embeddings (CE) can improve SI detection. Hence, this article proposes a method that combines CE with classification models generated by machine learning algorithms, to detect SI. The results obtained in the preliminary experiments with the proposed method presented pieces of evidence that the raised hypothesis is valid.

Palavras-chave: Neural networks, Text classification, Embeddings, BERT, Suicide ideation

Referências

Nayron Almeida, José Flávio, Breno Silva, Francisco Sousa, João Pedro Feitosa, Gerson Guimarães, and Luis Fernando Maia. 2018. Classificação de Risco de Suicídio Utilizando Análise deLinguagem Natural. (2018), 13–19

R.; CHISHMAN ALUíSIO, S.; CHECCHIA. [n. d.]. R. Brazilian portuguese liwc 2007. [link]

Gupta S. Sourirajan V. Belouali, A.2021. Acoustic and language analysis of speech for suicidal ideation among US veterans. BioData Mining) 14, 11 (2021), 1756–0381

Maria Tereza Camargo Biderman. 1996. Léxico e vocabulário fundamental. ALFA: Revista de Linguística 40 (1996)

Marouane Birjali, Abderrahim Beni-Hssane, and Mohammed Erritali. 2017. Machine Learning and Semantic Sentiment Analysis based Algorithms for Suicide Sentiment Prediction in Social Networks. Procedia Computer Science 113 (2017), 65–72. https://doi.org/10.1016/j.procs.2017.08.290 The 8th International Conference on Emerging Ubiquitous Systems and Pervasive Networks (EUSPN 2017) / The 7th International Conference on Current and Future Trends of Information and Communication Technologies in Healthcare (ICTH-2017) / Affiliated Workshops

Vinícius Cardoso, Antonio Silva, Roberta Sinoara, Solange Rezende, and Dario Calçada. 2019. Detecting Suicidal Ideation on Tweets. In Anais do XVI Encontro Nacional de Inteligência Artificial e Computacional (Salvador). SBC, Porto Alegre, RS, Brasil, 178–189. https://doi.org/10.5753/eniac.2019.9282

Gema Castillo-Sánchez, Gonçalo Marques, Enrique Dorronzoro, Octavio Rivera-Romero, Manuel Franco-Martín, De la Torre-Díez, 2020. Suicide risk assessment using machine learning and social networks: A scoping review. Journal of medical systems 44, 12 (2020), 1–15.

Conselho Federal de Medicina. 2014. Suicídio: informando para prevenir. Associação Brasileira de Psiquiatria) 14, 11 (2014), 9–11 [link].

Rodolpho da Silva Nascimento, Pedro Parreira, Gabriel dos Santos, and Gustavo Paiva Guedes. 2018. Identificando Sinais de Comportamento Depressivo em Redes Sociais. In Anais do VII Brazilian Workshop on Social Network Analysis and Mining (Natal). SBC, Porto Alegre, RS, Brasil. https://doi.org/10.5753/brasnam.2018.3597

Kelly Piacheski de Abreu, Maria Alice Dias da Silva Lima, Eglê Kohlrausch, and Joannie Fachinelli Soares. 2010. Comportamento suicida: fatores de risco e intervenções preventivas. Revista eletrônica de enfermagem 12, 1 (2010)

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 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). Association for Computational Linguistics, Minneapolis, Minnesota, 4171–4186. https://doi.org/10.18653/v1/N19-1423

Maite Gimenez, Javier Palanca, and Vicent Botti. 2020. Semantic-based padding in convolutional neural networks for improving the performance in natural language processing. A case of study in sentiment analysis. Neurocomputing 378 (2020), 315–323.

Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. RoBERTa: A Robustly Optimized BERT Pretraining Approach. CoRR abs/1907.11692 (2019). arxiv:1907.11692 http://arxiv.org/abs/1907.11692

Organização Mundial da Saúde OMS. 2019. Suicide Worldwide in 2019. [link].

Omar Oseguera, Alex Rinaldi, Joann Tuazon, and Albert C Cruz. 2017. Automatic quantification of the veracity of suicidal ideation in counseling transcripts. In International Conference on Human-Computer Interaction. Springer, 473–479

MARCIA Teresa Siebel, Bruna da Silva Santos, Líbia Miranda Moreira, and Viviane Silva Santos. 2019. A influência das redes sociais para o suicídio na adolescência. Revista Ciência (In) Cena 1, 8 (2019)

Fábio Souza, Rodrigo Nogueira, and Roberto Lotufo. 2019. Portuguese Named Entity Recognition using BERT-CRF. arXiv preprint arXiv:1909.10649 (2019). http://arxiv.org/abs/1909.10649

Michael Mesfin Tadesse, Hongfei Lin, Bo Xu, and Liang Yang. 2020. Detection of Suicide Ideation in Social Media Forums Using Deep Learning. Algorithms 13, 1 (2020). https://doi.org/10.3390/a13010007

Sulla-Torres J. Valeriano K., Condori-Larico A.2020. Detection of suicidal intent in Spanish language social networks using machine learning. BRAZILIAN WORKSHOP ON SOCIAL NETWORK ANALYSIS AND MINING (BRASNAM) 11, 4 (2020), 688–698.