Incorporating time-interval sequences in linear TV for next-item prediction

作者：

Highlights：

• Modeling user viewing habits in the TV domain as sequences of items.

• The user profile is represented by a vector of a 24-hour 15-minute slot views.

• Data shifting as a data augmentation method to predict any program at any time.

• Experiments were conducted with LSTM, Hidden Markov Models and a naive approach.

摘要

•Modeling user viewing habits in the TV domain as sequences of items.•The user profile is represented by a vector of a 24-hour 15-minute slot views.•Data shifting as a data augmentation method to predict any program at any time.•Experiments were conducted with LSTM, Hidden Markov Models and a naive approach.

论文关键词：Sequences,Recommender systems,TV domain

论文评审过程：Received 7 September 2020, Revised 20 September 2021, Accepted 21 November 2021, Available online 11 December 2021, Version of Record 21 December 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116284