This project aims to solve the problems of prisoners’ behavioral and psychological dynamics and concealment as well as to address their complex risk dimensions. Considering the lack of personalized and real-time risk assessment, this project aims to achieve full-cycle risk quantification, hierarchical and classified assessment, visual presentation and immediate supervision in order to realize the intelligent scheduling configuration of the disposal strategy.

Importance/impact

This project can scientifically classify and dispose of persons who were either released from prison or under community correction.  It can help social comprehensive management institutions and community correction institutions to carry out targeted social management and education.

Key contribution/commercial implication The information collection of prisoners in some pilot prisons and the construction of a multi-dimensional heterogeneous risk quantitative database have been completed;
The construction of the risk quantification system has been preliminarily completed.
Next steps
Settled in the pilot prison to improve and optimize the collection items in the database;
Use machine learning models to build a risk prediction model for prisoners Collaborators/partners
Shandong University
Qingdao Blue Ocean Information Industry Technology Research Institute
Beijing Thunisoft Information Technology Co.,Ltd. Team/contributors
PI: Zhuang Liu; Research Group: Yuhao Wu, Yuxia Zhang, Tianxiang Wang, Xiaoping Wu, Yijiao Liu