Prediction of Self-Harm Trends Using Machine Learning

DOI

10.2991/978-94-6463-471-6_43How to use a DOI?

Keywords

Self-harm; cross-lingual text classification; forecasting; nowcasting; and online social networks

Abstract

People hurt themselves by poisoning or hurting themselves in ways that cause injuries or death, even if they don't mean to. This is called self-harm. Self-harm not only hurts the people who do it, but it also hurts the income of the whole country. Self-harm is becoming more common, and studies have found a link between this and rapid growth of cities in developing countries and new technologies. It may be crucial for policymakers and public health professionals to forecast and anticipate national self-harm trends. But in some countries, it might be hard to get these kinds of past data or there might not be enough of it to make accurate predictions. This makes it harder to understand and predict the national self-harm landscape quickly. This essay suggests FAST, a system that will look at mental signs from a lot of social media data to predict trends of self-harm on a national level. These signs can be used as a stand-in for the mental health of the whole community and could be used to make it easier to predict trends in self-harm. These signals are combined into multivariate time series. Then, the time-delay embedding approach embeds these occurrences in time. Finally, several machine learning regressors are tested for future prediction. A Thailand case study found that 12 mental indications from tweets may predict self-harm-related mortality and injuries. The recommended technique predicted self-harm fatalities and injuries 43.56% and 36.48% better than ARIMA baseline. We believe our research is the first to utilize social media data to forecast and anticipate self-harm trends. Results not only help us figure out better ways to predict trends in selfharm, but they also lay groundwork for new social network-based apps that depend on being able to guess socioeconomic factors. We tried the Decision Tree algorithm and the Voting regressor, which are the best machine learning algorithms. These algorithms gave us lower MAE errors than other algorithms.

Copyright

© 2024 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - C. Siva Kumar
AU  - P. Lakshmi Sagar
AU  - Patnam Venkataiah
AU  - Setti Partha Saradhi
AU  - Annam Mohan Kumar
AU  - Velagaleti Bhavan
PY  - 2024
DA  - 2024/07/30
TI  - Prediction of Self-Harm Trends Using Machine Learning
BT  - Proceedings of the International Conference on Computational Innovations and Emerging Trends (ICCIET- 2024)
PB  - Atlantis Press
SP  - 441
EP  - 452
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-471-6_43
DO  - 10.2991/978-94-6463-471-6_43
ID  - Kumar2024
ER  -