Identifying suicidal emotions on social media through transformer-based deep learning

Dheeraj, Kodati; Ramakrishnudu, Tene

doi:10.1007/s10489-022-04060-8

Cited by 7 publications

(2 citation statements)

References 70 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the process of predicting online public opinion crises, it is necessary to further analyze the large amount of relevant information studied and collected, and establish a scientific, reasonable, Frontiers in Educational Psychology effective, practical, operable, and objective model that can reveal the relationship between regular features and phenomena. On the basis of qualitative prediction of crisis events, a model was established to quantitatively analyze online public opinion crises [17][18]. This method is mainly used to study the probability, intensity, and development trend of events.…”

Section: Experimental Planmentioning

confidence: 99%

Qualitative Prediction and Quantitative Analysis of Events in Educational Network Public Opinion Crisis: a Deep Learning Approach

2023

JEP

View full text Add to dashboard Cite

The crisis of educational network public opinion refers to issues such as public dissatisfaction and trust crisis caused by unexpected events in the field of education. The frequency and scope of the impact of educational network public opinion crisis are gradually expanding, bringing huge challenges to educational management departments and institutions. This article aimed to explore how to use deep learning technology to qualitatively predict and quantitatively analyze educational network public opinion crises. This article adopted experimental and comparative methods to qualitatively and quantitatively analyze the crisis of public opinion in educational networks, and obtained the characteristics of several deep learning methods. Experimental data showed that under the framework of convolutional neural network (CNN), the accuracy of public opinion crisis prediction in the sigmoid function was 99.9%; the loss rate was 0.34%, and the running time was 65 seconds.

show abstract

Section: Experimental Planmentioning

confidence: 99%

Qualitative Prediction and Quantitative Analysis of Events in Educational Network Public Opinion Crisis: a Deep Learning Approach

2023

JEP

View full text Add to dashboard Cite

show abstract

“…In recent years, the proliferation of digital platforms and social media has provided an unprecedented opportunity to capture and analyze large-scale data related to mental health [2] [3]. Machine learning and Natural Language Processing (NLP) techniques have shown promise in detecting linguistic patterns and indicators of suicidal ideation in diverse textbased data sources, such as social media posts, online forums, and electronic health records [4], [5], [6], [7].…”

Section: Introductionmentioning

confidence: 99%

Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language Models

Ghanadian,

Nejadgholi,

Osman

2024

IEEE Access

View full text Add to dashboard Cite

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 F1scores 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.

show abstract

Research on automatic error-checking in English short text translation by a neural network algorithm

Guo

2024

Artif Life Robotics

View full text Add to dashboard Cite

Identifying suicidal emotions on social media through transformer-based deep learning

Cited by 7 publications

References 70 publications

Qualitative Prediction and Quantitative Analysis of Events in Educational Network Public Opinion Crisis: a Deep Learning Approach

Qualitative Prediction and Quantitative Analysis of Events in Educational Network Public Opinion Crisis: a Deep Learning Approach

Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language Models

Research on automatic error-checking in English short text translation by a neural network algorithm

Contact Info

Product

Resources

About