71 / 2023-06-30 20:59:29

Research on Mining Users’ Daily Requirements Considering Variable-order Dependencies Behavior Analysis Based on Higher-order Dependency Networks

higher-order dependency networks,community detection,random walk

大数据与复杂系统 > 2.复杂系统与高阶网络

In the era of the popularization of the Internet of Things(IOT), analyzing people’s daily life behavior through the information collected by sensors is an important method to tap potential daily needs. The network method is an important means to analyze the relationship between people’s daily behaviors, while the mainstream first-order network(FON) method ignores the high-order dependencies between daily behaviors. Higher-order dependency networks(HON) can more accurately mine the requirements by considering higher-order dependencies and historical path traceability. Firstly, this paper adopts indoor daily behavior sequences obtained by video behavior detection, extracts higher-order dependency rules from behavior sequences, and rewires a HON for behavior. Secondly, HON is used for the random walk algorithm. On this basis, researches on vital node identification and community detection are carried out. Finally, research results on behavioral datasets show that, compared with FON, HON can significantly improve the accuracy of random walks, improve the identification of important nodes, and make community membership results more diverse. The purpose of this research is to improve the effect of personalized recommendation and requirement mining. Meanwhile, this method can also be used to improve the accuracy and interpretability of indoor construction of user portraits, personalized recommendations, and hidden danger investigation.