Treffer: Twitter SOS: Predicting suicide tendency in twitter data using ML.

The content analysis of text to identify suicidal tendencies and different types. It outlines the development of a sentence classifier using a neural network, implemented through various Python machine learning libraries. The focus is on addressing the pressing issue of teenage suicide and the proliferation of "groups of death" on social networks, while seeking effective ways to combat the promotion of suicide among minors. The presentation also explores the analysis of existing information pertaining to these "groups of death" and their online dissemination. Additionally, we evaluate the performance of multiple classification algorithms, determining the most accurate one for identifying patterns of suicidal conversations. Our findings offer valuable insights into the early detection and prevention of suicide-related content. Beginning with data preprocessing, you utilized the preprocess_ kgptalkie library for tasks like text cleaning and contraction expansion. The script proceeds to employ a TF-IDF vectorizer for feature extraction, followed by the training of multiple classifiers including Linear SVC, Logistic Regression, Decision Tree, Random Forest, AdaBoost, Gradient Boosting, and Bagging. The conclusion rightly identifies Support Vector Machine (SVM) as the optimal model based on its superior accuracy and a detailed examination of the confusion matrix. Real-time predictions are showcased using the trained SVM model, demonstrating its applicability for deployment in practical scenarios. Additionally, a mention of an abstract for a report on suicidal ideation detection in tweets is made, though the actual content is not provided. This script serves as a robust framework for developing and deploying a machine learning model for the critical task of identifying potentially suicidal content in social media posts. [ABSTRACT FROM AUTHOR]

Copyright of AIP Conference Proceedings is the property of American Institute of Physics and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)