DNCP: An attention-based deep learning approach enhanced with attractiveness and timeliness of News for online news click prediction
Abstract: Highlights•This study defines the concepts of attractiveness and timeliness of news and proposes new methods for representing them.•We propose a deep news click prediction (DNCP) method to integrate attractiveness, timeliness, and textual and meta features of news for news click prediction.•We enhance the attention representation of a news article based on the attractiveness and timeliness of individual terms contained in the article.•This study provides empirical evidence about the effectiveness of DNCP in news click prediction, providing novel technical and practical insights.
Loading