Mini-review of street crime prediction and classification methods

• Main Authors: Nurul Farhana Mohamad Zamri, Nooritawati Md Tahir, Megat Syahirul Amin Megat Ali, Nur Dalila Khirul Ashar, Al-misreb, Ali Abd
• Format: Article
• Language: English
• Published: Penerbit Universiti Kebangsaan Malaysia 2021
• Online Access: http://journalarticle.ukm.my/18749/1/02.pdf; http://journalarticle.ukm.my/18749/; https://www.ukm.my/jkukm/volume-333-2021/
• Institution: Universiti Kebangsaan Malaysia

Crime rates are one of the biggest problems in today’s modern society, especially in urban cities. Various techniques on crime prediction and detection have been developed by previous researchers in reducing the crime rates that keep increasing throughout the year as well as to assist the government authorities in combating crimes. These include studies on forecasting crime activities based on both primary and secondary data that include numerical data, statistics, video, and images related to various categories of crimes. Thus, in this study, a mini-review is conducted related to the database used as well as methods that have been developed by previous researches related to crime classification, crime analysis and forecasting of crime or crime prediction. Further, a new technique will be proposed in the detection of crime activities. The proposed technique involves evaluation and validation of several Deep Learning (DL) specifically the Convolutional Neural Network (CNN) along with the type of database to be used specifically for street crime detection that focuses on snatch theft.