Recommending Electricity Plan: A Significant Nearest Neighbors Collaborative Filtering Method
Ning Ye, Xiying Gao, Guan Yan · 2021
With the development of electricity market trading, building owners will possess the freedom to choose their preferred electricity retailers. This paper inspects the application of recommender system. Different from the traditional KNN (K-nearest neighbors)-based collaborative filtering method, we chooses the SNN (significant nearest neighbors)-based collaborative filtering method. The users with top relative similarity index values with respect to the target user are called SNN. This method effectively reduces the computational cost and improves the performance of the personalized recommendation system. By using weekly operation duration times of some typical electrical appliance data, the recommendation system will rate a large number of different plans, so that the homeowner can choose a more suitable electricity package. Different tests can evaluate the performance of the method. The instructions of the proposed method on electricity retailer helps to improve the competitive operation.