Deep Learning Algorithms for Suicide Prediction Based On Bilateral Long-Term Memory Using Social Media Behaviour Dataset
Author(s):
R. Parthiban, S. Siva Selvi, G.Sivakumar
Keywords:
Suicide rates, TF-IDF, deep learning, Bilateral Long-Term Memory (BiLSTM), Deep Convolutional Neural Networks (DCNN), sentiment analysis, features.
Abstract
Suicide rates are one of the most severe issues in the world. The rate of suicides is rising at a rapid rate over time. To predict the causes of suicide in India, this work uses deep learning algorithms to identify the underlying causes of suicides. However, identifying and understanding patterns of suicidal ideation can be difficult. Hence, there is a need to develop deep learning systems that can automatically and proactively detect suicidal thoughts and sudden changes in user behaviour by analyzing users' social media posts. We suggest an experimental research-based approach to create a method to detect suicidal ideation using word-encoding techniques like TF-IDF and Word2Vec and deep learning for classification using the publicly available Reddit dataset. Utilize a Bilateral Long-Term Memory (BiLSTM) model to categorize social media messages as suicidal. Initially collecting the data from the standard repository and then reducing the missing and irrelevant dataset values in the preprocessing stage. The second stage is extracting the features based on the threshold weights using Deep Convolutional Neural Networks (DCNN) are often used for feature extraction and data dimensionality reduction. DCNN can extract complex features, describe images in more detail, learn task-specific features, and be more efficient. Bilateral Long-Term Memory (BiLSTM) is widely used in text mining and sentiment analysis. On the other hand, it uses BiLSTM to extract the most essential and reliable features for classification automatically. To estimate the standard metrics such as precision, accuracy, Recall, and F1 score to evaluate model performance.
Article Details
Unique Paper ID: 159184

Publication Volume & Issue: Volume 9, Issue 11

Page(s): 930 - 935
Article Preview & Download


Share This Article

Join our RMS



Conference Alert

NCSEM 2024

National Conference on Sustainable Engineering and Management - 2024

Last Date: 15th March 2024

Latest Publication

Call For Paper

Volume 11 Issue 1

Last Date for paper submitting for Latest Issue is 25 June 2024

About Us

IJIRT.org enables door in research by providing high quality research articles in open access market.

Send us any query related to your research on editor@ijirt.org

Social Media

Google Verified Reviews

Contact Details

Telephone:6351679790
Email: editor@ijirt.org
Website: ijirt.org

Policies

Important Links

Browse Rsearch papers