A hashtag is defined as any word attached to the prefix character '#' that is used in online social network services (SNS) such as Facebook, Twitter, and Instagram. With the growth of online social networks, hashtags are commonly used to summarize the content of a user's post and attract the attention of followers. The recommendation of proper hashtags is a highly interesting and useful task in the age of social media.

We introduce the novel benchmark for image hashtag recommendation, titled HARRISON, or the HAshtag Recommendation for Real-world Images in SOcial Networks. The HARRISON dataset is a realistic dataset, composed of 57,383 photos from Instagram and an average of 4.5 associated hashtags for each photo (minimum 1 and maximum 10). The ground truth hashstags for each image are made up of the 1,000 most frequently used hashtags, encoded with numbers based on frequency ranking results.

Overview