Recommender System for Crime Against Women

Crime against women is a major issue for any country, state or city. The meaning of the crime against women is defined as mental physical brutality associated with women. The research is primarily focused on recommending a model for classification and prediction of offenders under sexual crime against women. To start with, the study first focused on identification of factors which affect crime or offences against women from the existing studies. Then significant impact of each factor has been checked by applying regression analysis on data collected from respondents upon identified factors namely Lack of safe spaces for women & girls, Conflict and tension within intimate partner, relationship or marriage, Gender discrimination, Inappropriate presentation of women characters on TV or silver screens, Misuse of Electronic and social media, Poor upbringing, Poverty, Lack of knowledge about Acts, Lack of implementations of government initiatives, Lack of punishment and Male Ego.  

As a result it has been found that all eleven identified factors have their impact on crime against women because there is a significance relationship of factors with crime against women. Further a well-structured conceptual model has been developed after applying a mathematical approach namely Interpretive structural modelling (ISM). Interpretive structural modelling technique or procedure starts to find the interrelationship of various factors related to the issue. It is an effective process or mathematical approach to solve complex problems. Based on brainstorming sessions and experts opinion, a matrix has been depicted that shows the contextual relationship between all validated factors affecting crime against women.  

This matrix is called structural self-interaction matrix (SSIM), an initial step of ISM technique. Further, a new matrix has been constructed called "Reachability Matrix (RM)", derived with the help of SSIM. The purpose of reachability matrix is to find all direct or indirect links between the factors, which indicates 'how one factor is leading to another one by direct or indirect way'. After building of the final ISM model, MICMAC analysis has been done which classified factors into four clusters namely strategic, operational and dependent factors as per their driving and dependence power.