δ-CHUCPM: A δ-closed high utility co-location pattern miner
Vanha Tran, Cao Dai Pham, Thanhcong Do, Hoangnam Pham · 2023
High utility co-location patterns (HUCPs) are more meaningful than traditional prevalent co-location patterns (PCPs) since HUCPs consider the utility of instances that reflects the significance of the instances in reality. However, the mining result normally contains too many HUCPs, this makes HUCPs difficult for users to understand, absorb and apply. This demonstration presents a δ-closed high utility co-location pattern miner (δ-CHUCPM), a mining system that removes redundant HUCPs and presents a concise representation of the mining result to users. We achieve the removing redundant HUCP objective by allowing a tolerance between the utility of a pattern and the utility of its superset. If the tolerance is not larger than a user-specified small value δ, the pattern is removed. Users can freely set different δ tolerance values according to their own application scenarios and requirements to control the number of HUCPs. δ-CHUCPM can provide users with a more understandable result and help them make informed decisions.