Prediction of Crime Occurrence in case of Scarcity of Labeled Data

Abstract

In line with technological developments, machine learning/data mining studies have significantly scaled up in crime analysis. The prediction of crime occurrences, the detection of the spatial/temporal distribution of the criminal cases, forecasting the type of crime are some of these study areas. By taking crime data resulting from a substantial increase in crime rates into consideration, unlabeled data can be utilized to enhance exploring the patterns of crime for future events or to make crime-related predictions easily. Therefore, in this study, active learning, selflearning, and random sampling techniques are applied to predict the outcome of criminal searches in England using the police data of 2019. According to the experimental analysis, active learning outperforms its counterparts using its entropy-based smart selection strategy data in case there is little labeled data.

Analysis of Legal Issues of the Crime of Endangering Public Safety Based on Data Mining Algorithm

“…For the understanding of public danger, we should not only take into account the purely objective and physical dangers, but also pay attention to the psychological factors, that is, to identify the public danger concept by capturing the insecurity of ordinary people, which requires data mining algorithms to carry out behavior classification processing [11]. Regarding the understanding of data mining for dangerous data processing, the literature [12] pointed out that there are two different understandings: one is the behavioral risk theory that the behavior produces the danger of infringing the protected interests, and the other is that the actor is antagonistic, that is, the theory of danger of dangerous actors in society. With the development of the objectivist criminal law theory, the behavioral risk theory gradually replaces the perpetrator's risk theory.…”