Open AccessResearch on fused sorting based on logical regression in news recommendation system
Author(s) -
Ling Xing,
Xiao Feng,
Haiming Chen,
Ying Wang,
Yue Zhang
Publication year - 2020
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/510/6/062029
Subject(s) - sorting , computer science , recall , set (abstract data type) , information retrieval , precision and recall , data mining , sorting algorithm , data set , recommender system , channel (broadcasting) , artificial intelligence , sort , algorithm , computer network , linguistics , philosophy , programming language
A fusion sorting model based on logical regression is used to solve the problem of information overload in the news recommendation system and meet the personalized needs of users. First, the fusion sorting model is trained based on the article data and user data, and the original candidate news set is obtained by multi-channel recall method. Then the original candidate set is filtered by the fusion sorting model, and the recommendation list is generated. Finally, the experiment is carried out based on the Wuli News Recommendation System. Under the evaluation indexes of precision, recall and ILS, the fusion sorting model adopted significantly improves the recommendation ability of the system, and achieves the goal of making multiple single recommendation strategies make up for each other’s shortcomings.