A ListNet-based Combination of Lexical and Semantic Features to Homogenize Folksonomies in Online Social Networks
Fahd Kalloubi, El Habib Nfaoui · 2023
The popularity of online social networks has grown tremendously in recent years, with over 4.6 billion active users in 2022. Hashtags are commonly used by social media users to categorize their posts and connect with like-minded individuals. However, with the sheer volume of hashtags being used, it can be challenging and time-consuming to select the most appropriate ones. Therefore, there is a need for a system that can help users choose suitable hashtags. In this paper, we propose a learning-to-rank-based approach that considers both lexical and semantic features to recommend hashtags on microblogging platforms. We also investigate the impact of combining different ranking strategies on hashtag suggestions. Our experiments, which used a large dataset from Twitter, demonstrate that our approach outperforms both lexical-based and semantic-based methods.