Which, Who and How: Detecting fraudulent sale accumulation behavior from multi-dimensional sparse data
作者:
Highlights:
• Detecting sale accumulation behavior from multi-dimensional sparse data.
• Proposing a new concept of partially ordered lattice of the data warehouse schema.
• Proposing sale accumulation mining algorithms and sale accumulation pattern classification methods.
• Handling the problems of data sparsity and the coexistence of the normal and abnormal high sale volumes.
摘要
•Detecting sale accumulation behavior from multi-dimensional sparse data.•Proposing a new concept of partially ordered lattice of the data warehouse schema.•Proposing sale accumulation mining algorithms and sale accumulation pattern classification methods.•Handling the problems of data sparsity and the coexistence of the normal and abnormal high sale volumes.
论文关键词:Commercial fraud detection,Sale accumulation,Sale accumulation pattern,Partially ordered lattice,Feature extraction
论文评审过程:Received 25 January 2021, Revised 20 June 2021, Accepted 24 June 2021, Available online 26 June 2021, Version of Record 30 June 2021.
论文官网地址:https://doi.org/10.1016/j.knosys.2021.107260