Geo-Spatial Crime Analysis Using Newsfeed Data in Indian Context

2. Literature Review

Historical crime data can be used to identify high crime areas and plan resources optimally. Predictive policing using data enables law enforcement authorities to take proactive decisions to improve response time to crime incidents (Angers, Biswas & Maiti, 2016). Knowledge acquired from the data mining techniques can be used to helping find criminals faster thereby reducing crime rate. crime prediction, a subtask of crime analysis, considers all the past crime records, classifies the crime categories and predicts the future crime. Crime prediction using pattern and association rule mining determines the chances of performing crime by the same criminal.

Research by urban activist Jane Jacobs (2012) emphasizes that natural surveillance, i.e., the presence of high density of visitors and high diversity enhance the safety of the target area and in effect reduces crime. With the help of data-driven and place-centric approach, it is possible to determine whether a particular geographic area can be identified as a future crime area proposed by X. Wang et al. (2012). Crime models based on spatial analysis created from newsfeed proposed by Jayaweera et al. (2015) can help us in understanding criminal activity better. Also, this anonymized data has the advantage of limited to no privacy risks. Combining anonymized and aggregated user data with demographics can help in identifying whether specific locations are more prone to future crimes or not.