Determining domestic violence against women using machine learning methods: The case of Türkiye

Abstract

BACKGROUND: Domestic violence against women is a pervasive issue globally, representing a severe violation of human rights and a significant public health concern. The hidden nature of such violence and its frequent underreporting make it a critical area for research. Recent advancements in artificial intelligence offer new avenues for identifying and predicting instances of domestic violence through machine learning (ML) algorithms. AIM: This study aimed to determine the frequency and risk factors of domestic violence against women using ML methods. METHODS: With a cross-sectional design, this research was conducted with 630 married women between December 2023 and February 2024. Data were obtained using the 'Demographic Information Form' and the 'HITS Domestic Violence Scale'. Data analysis used six ML algorithms (decision tree, random forest, support vector machine [SVM], logistic regression [LR], Naive Bayes and k-nearest neighbour). RESULTS: In our study, the rate of women experiencing violence was determined to be 11%, while the duration of marriage, number of children and level of education were identified as significant risk factors. Threat, insult and injury were common risk factors in all algorithms. SVM and LR algorithms were effective models in predicting violence with a 100% accuracy rate. All ML algorithms' sensitivity ranged from 91.12% to 100%, while specificity ranged from 85% to 100%. CONCLUSION: The findings of our study demonstrate that ML algorithms have high accuracy rates in determining the frequency and risk factors of domestic violence against women, indicating that they can be used safely.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 10 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.