Predicting time and location of future crimes with recommendation methods
作者:
Highlights:
• Crime prediction can be formulated as a recommendation problem.
• Recommendation methods mitigate the sparsity problem in crime prediction.
• Biased matrix factorization is designed to incorporate social context.
• CBMF outperforms state-of-the-art recommendation and non-recommendation methods.
摘要
•Crime prediction can be formulated as a recommendation problem.•Recommendation methods mitigate the sparsity problem in crime prediction.•Biased matrix factorization is designed to incorporate social context.•CBMF outperforms state-of-the-art recommendation and non-recommendation methods.
论文关键词:Crime prediction,Recommendation system,Social media
论文评审过程:Received 25 February 2020, Revised 28 September 2020, Accepted 1 October 2020, Available online 5 October 2020, Version of Record 8 October 2020.
论文官网地址:https://doi.org/10.1016/j.knosys.2020.106503