Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language Models
Suicidal ideation detection is a vital research area that holds great potential for improving mental health support systems. However, the sensitivity surrounding suicide-related data poses challenges in accessing large-scale, annotated datasets necessary for training effective machine learning model...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2024-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10413447/ |
_version_ | 1797335521608859648 |
---|---|
author | Hamideh Ghanadian Isar Nejadgholi Hussein Al Osman |
author_facet | Hamideh Ghanadian Isar Nejadgholi Hussein Al Osman |
author_sort | Hamideh Ghanadian |
collection | DOAJ |
description | Suicidal ideation detection is a vital research area that holds great potential for improving mental health support systems. However, the sensitivity surrounding suicide-related data poses challenges in accessing large-scale, annotated datasets necessary for training effective machine learning models. To address this limitation, we introduce an innovative strategy that leverages the capabilities of generative AI models, such as ChatGPT, Flan-T5, and Llama, to create synthetic data for suicidal ideation detection. Our data generation approach is grounded in social factors extracted from psychology literature and aims to ensure coverage of essential information related to suicidal ideation. In our study, we benchmarked against state-of-the-art NLP classification models, specifically, those centered around the BERT family structures. When trained on the real-world dataset, UMD, these conventional models tend to yield F1-scores ranging from 0.75 to 0.87. Our synthetic data-driven method, informed by social factors, offers consistent F1-scores of 0.82 for both models, suggesting that the richness of topics in synthetic data can bridge the performance gap across different model complexities. Most impressively, when we combined a mere 30% of the UMD dataset with our synthetic data, we witnessed a substantial increase in performance, achieving an F1-score of 0.88 on the UMD test set. Such results underscore the cost-effectiveness and potential of our approach in confronting major challenges in the field, such as data scarcity and the quest for diversity in data representation. |
first_indexed | 2024-03-08T08:39:29Z |
format | Article |
id | doaj.art-9b409c1d5d9d488e9cafe8e1172c76ca |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-03-08T08:39:29Z |
publishDate | 2024-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-9b409c1d5d9d488e9cafe8e1172c76ca2024-02-02T00:04:22ZengIEEEIEEE Access2169-35362024-01-0112143501436310.1109/ACCESS.2024.335820610413447Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language ModelsHamideh Ghanadian0https://orcid.org/0000-0002-5203-3504Isar Nejadgholi1Hussein Al Osman2https://orcid.org/0000-0002-7189-5644Department of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, ON, CanadaNational Research Council Canada, Ottawa, ON, CanadaDepartment of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, ON, CanadaSuicidal ideation detection is a vital research area that holds great potential for improving mental health support systems. However, the sensitivity surrounding suicide-related data poses challenges in accessing large-scale, annotated datasets necessary for training effective machine learning models. To address this limitation, we introduce an innovative strategy that leverages the capabilities of generative AI models, such as ChatGPT, Flan-T5, and Llama, to create synthetic data for suicidal ideation detection. Our data generation approach is grounded in social factors extracted from psychology literature and aims to ensure coverage of essential information related to suicidal ideation. In our study, we benchmarked against state-of-the-art NLP classification models, specifically, those centered around the BERT family structures. When trained on the real-world dataset, UMD, these conventional models tend to yield F1-scores ranging from 0.75 to 0.87. Our synthetic data-driven method, informed by social factors, offers consistent F1-scores of 0.82 for both models, suggesting that the richness of topics in synthetic data can bridge the performance gap across different model complexities. Most impressively, when we combined a mere 30% of the UMD dataset with our synthetic data, we witnessed a substantial increase in performance, achieving an F1-score of 0.88 on the UMD test set. Such results underscore the cost-effectiveness and potential of our approach in confronting major challenges in the field, such as data scarcity and the quest for diversity in data representation.https://ieeexplore.ieee.org/document/10413447/Artificial intelligencedeep learninglarge language modelssuicide detectionsynthetic data generationtransformer based models |
spellingShingle | Hamideh Ghanadian Isar Nejadgholi Hussein Al Osman Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language Models IEEE Access Artificial intelligence deep learning large language models suicide detection synthetic data generation transformer based models |
title | Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language Models |
title_full | Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language Models |
title_fullStr | Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language Models |
title_full_unstemmed | Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language Models |
title_short | Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language Models |
title_sort | socially aware synthetic data generation for suicidal ideation detection using large language models |
topic | Artificial intelligence deep learning large language models suicide detection synthetic data generation transformer based models |
url | https://ieeexplore.ieee.org/document/10413447/ |
work_keys_str_mv | AT hamidehghanadian sociallyawaresyntheticdatagenerationforsuicidalideationdetectionusinglargelanguagemodels AT isarnejadgholi sociallyawaresyntheticdatagenerationforsuicidalideationdetectionusinglargelanguagemodels AT husseinalosman sociallyawaresyntheticdatagenerationforsuicidalideationdetectionusinglargelanguagemodels |