Advanced deep learning and large language models for suicide ideation detection on social media

被引:1
作者
Qorich, Mohammed [1 ]
El Ouazzani, Rajae [1 ]
机构
[1] Moulay Ismail Univ Meknes, Sch Technol, IMAGE Lab, ISNET Team, Meknes, Morocco
关键词
Bidirectional long short-term memory (BiLSTM); Convolutional neural network (CNN); Large language models (LLMs); Mental health; Natural language processing (NLP); Suicide ideation (SI); Text classification; Triple word embedding;
D O I
10.1007/s13748-024-00326-z
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Recently, suicide ideations represent a worldwide health concern and pose many anticipation challenges. Actually, the prevalence of expressing self-destructive thoughts especially on forums and social media requires effective monitoring for suicide prevention, and early intervention. Meanwhile, deep learning techniques and Large Language Models (LLMs) have emerged as promising tools in diverse Natural Language Processing (NLP) tasks, including sentiment analysis and text classification. In this paper, we propose a deep learning model incorporating triple models of word embeddings, as well as various fine-tuned LLMs, to identify suicidal thoughts in Reddit posts. In effect, we implemented a Bidirectional Long Short-Term Memory (BiLSTM), and a Convolutional Neural Network (CNN) model to categorize posts associated with non-suicidal and suicidal thoughts. Besides, through the combination of Word2Vec, FastText and GloVe embeddings, our models learn intricate patterns and prevalent nuances in suicide-related language. Furthermore, we employed a merged version of CNN and BiLSTM models, entitled C-BiLSTM, and several LLMs, including pre-trained Bidirectional Encoder Representations from Transformers (BERT) models and a Generative Pre-training Transformer (GPT) model. The analysis of all our proposed models shows that our C-BiLSTM model with triple word embedding and our GPT model got the best performance compared to deep learning and LLMs baseline models, reaching accuracies of 94.5% and 97.69%, respectively. In fact, our best model's capacity to extract meaningful interdependencies among words significantly promotes its classification performance. This analysis contributes to a deeper understanding of the psychological factors and linguistic markers indicative of suicidal thoughts, thereby informing future research and intervention strategies.
引用
收藏
页码:135 / 147
页数:13
相关论文
共 51 条
  • [1] Detecting and Analyzing Suicidal Ideation on Social Media Using Deep Learning and Machine Learning Models
    Aldhyani, Theyazn H. H.
    Alsubari, Saleh Nagi
    Alshebami, Ali Saleh
    Alkahtani, Hasan
    Ahmed, Zeyad A. T.
    [J]. INTERNATIONAL JOURNAL OF ENVIRONMENTAL RESEARCH AND PUBLIC HEALTH, 2022, 19 (19)
  • [2] Alsentzer E, 2019, P 2 CLIN NAT LANG PR, P72, DOI DOI 10.18653/V1/W19-1909
  • [3] Ananthakrishnan G., 2022, 3 IEEE 2022 INT C CO, P322, DOI DOI 10.1109/ICCCIS56430.2022.10037677
  • [4] Asim MN, 2017, INT J ADV COMPUT SC, V8, P369, DOI 10.14569/IJACSA.2017.081048
  • [5] Text Classification to Inform Suicide Risk Assessment in Electronic Health Records
    Bittar, Andre
    Velupillai, Sumithra
    Roberts, Angus
    Dutta, Rina
    [J]. MEDINFO 2019: HEALTH AND WELLBEING E-NETWORKS FOR ALL, 2019, 264 : 40 - 44
  • [6] Cao L, 2019, 2019 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING AND THE 9TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING (EMNLP-IJCNLP 2019), P1718
  • [7] Identification of suicidal behavior among psychiatrically hospitalized adolescents using natural language processing and machine learning of electronic health records
    Carson, Nicholas J.
    Mullin, Brian
    Sanchez, Maria Jose
    Lu, Frederick
    Yang, Kelly
    Menezes, Michelle
    Le Cook, Benjamin
    [J]. PLOS ONE, 2019, 14 (02):
  • [8] Chadha A., 2022, P INT C TECHN ADV CO, P685, DOI [10.1109/ICTACS56270.2022.9988722, DOI 10.1109/ICTACS56270.2022.9988722]
  • [9] A Hybrid Deep Learning Model Using Grid Search and Cross-Validation for Effective Classification and Prediction of Suicidal Ideation from Social Network Data
    Chadha, Akshma
    Kaushik, Baijnath
    [J]. NEW GENERATION COMPUTING, 2022, 40 (04) : 889 - 914
  • [10] A Survey on Prediction of Suicidal Ideation Using Machine and Ensemble Learning
    Chadha, Akshma
    Kaushik, Baijnath
    [J]. COMPUTER JOURNAL, 2021, 64 (11) : 1617 - 1632