Study of Violence Against Women and its Characteristics Using Application of Data Mining Techniques

Abstract

The Internet serves as a vast source of information, offering diverse data that can be gathered and analysed to create an extensive repository. Articles addressing the crucial issue of Violence Against Women (VAW) published online significantly enhance our understanding of this subject. In this study, we employed web scraping to collect VAW-related news, processed the data using a feature selection model to create a comma-separated dataset, and applied text mining techniques for comprehensive analysis. This included exploratory analysis and Topic Modelling to uncover latent topics. We also utilized classification algorithms such as Naive Bayes, Random Forest, Support Vector Machine (SVM), AdaBoost, and Artificial Neural Networks (ANN) to categorize the types of violence physical, psychological, sexual, or a combination. By integrating these techniques, our study provides a nuanced understanding of VAW, revealing patterns and trends that can inform targeted interventions and support mechanisms.