Twitter and other social media microblogging websites have gained popularity in recent years—and the large volume of information can be overwhelming for both novice and experienced users of these platforms. The hashtag emerged as a way for Twitter users to group and navigate through topics, by attaching a word or code to the end of the message, with the word or phrase being led by the “#” symbol. These keywords allowed users to group and receive news and messages that were relevant to that topic.
But the role of the hashtag may be more complex than the bookmarking of content. Hashtags can serve as a symbol of community, allowing users to virtually converse directly on the Twitter platform without the need for an external chat room. The dual role of the hashtag is to allow users to label and curate content related to a topic, but also become a member of a virtual community. “By creating a hashtag, a user either invents and shares a new bookmark (of content), or initializes and spreads a coat of arms (of a community), or both, ” write researchers, Lei Yang, Tao Sun, Ming Zhang, and Qiaozhu Mei at the University of Michigan’s School of Information and Peking University’s School of EECS. Generally, social tags serve two purposes: organizational and social.
Yang et. al. conducted a systematic empirical analysis of how hashtag’s dual role affects adoption of the hashtag itself. In a piece titled We Know What @You #Tag: Does the Dual Role Affect Hashtag Adoption?, they used the following measures related to content tagging: relevance and preference; measures related to joining a community: prestige and influence. Other measures include popularity, hashtag length, freshness, degree and activeness. Datasets for the research were drawn from political users from March 2007 to December 2010, as well as a random sample of 5 percent of roughly 19 million users, 49 million unique hashtags and 476 million retweets.
The findings suggest that hashtags function both as a way to tag content and a symbol of membership in a community, proving the effectiveness of the dual role where the content and community measures strongly correlate to hashtag use on Twitter. With these measures as features, a machine-learning model can predict the future adoption of hashtags that a user has never used before.