首页 > English > Home > Current Issue > 正文

13 Research on long short-term preference sequence recommendation algorithm based on LSTM

日期:2024-10-17 18:16:58 点击:

ZHAO Lulu ,  ZHAO Yuhong

(Digital and Intelligent Industry School ,  Inner Mongolia University  of Science and Technology ,  Baotou 014010 ,  China)


Abstract :  Dynamic  time series is a key feature of many modern recommendation systems .   Its primary aim is to capture the “context” of user activities based on their most recent actions . However ,  most LSTM-based  sequence models only consider the user ’s short-term interests ,  neglecting their long-term interests .   To enhance the performance of sequence recommendations ,  a Long Short-term Preference Recommendation Based on LSTM  (LLSPRec)  method is proposed here .   This method models the user ’s time series with LSTM ,  aggre- gates relevant feature information between sequences to obtain the user ’s recent preferences ,  and models the  distance between the user and candidate items using distance metric learning to capture the user ’s long-term preferences ,  dynamically integrating the user ’s long- term and short-term preferences according to their intentions ,  thereby accurately describing user interests and improving the diversity of recommendation results .

Key words :  LSTM ;  metric learning;  dynamic time  series;  sequence recommendation;  element correlation


地址:内蒙古包头市昆都仑区阿尔丁大街7号 邮编:014010 电话:0472-5951610或0472-5953910 Email:cky@imust.edu.cn nkdxb@imust.edu.cn

版权所有:内蒙古科技大学学报编辑部(©2013)