History-Related Content Recommendation in Social Networks Based on User's Interest Estimation
Kouki Yoshida, Adam Jatowt, Yasunobu Sumikawa · 2023
Knowing and analyzing the events of the past is important as it allows us to understand the formation process of modern society. Twitter users are able to acquire and share information of their interests including also content on history. We propose an algorithm to recommend history-related content according to the interests of Twitter users. Our approach analyzes users' past tweets to determine their interests by modeling the forgetting curve of memory. It determines the users' current interests by considering the time passage and the repetition of posted content on particular topics.