Interactive Hashtag Recommendation System

With the progressive advance of Internet technologies, more and more users share their lives by posting tweets on social media platforms like Twitter. These tweets use hashtags as links to constitute discussion topics on social media. However, since most users are not used to using hashtags, a large number of tweets cannot be related to corresponding topics. To solve this problem, we propose an interactive hashtag recommendation system, which predicts the topic of an input tweet and interactively recommends relevant hashtags in different phases of writing tweets. When users use recommended hashtags, the input tweet can be related to the corresponding topic by using hashtags. Tweets on social media can form many discussion topics through hashtags. As such, the system can help to build consensus about hashtags on social media. We conducted user experiments to verify the usability of the implemented recommendation system. The experimental results and user feedback reveal that this interactive hashtag recommendation system can provide accurate hashtags relevant to the corresponding topic.