Actively Learning to Attract Followers on Twitter
This addresses the challenge of actively learning follower acquisition strategies for Twitter users, though it is incremental as it builds on existing bandit methods applied to a new social media context.
The study tackled the problem of learning to acquire followers on Twitter through normative user behavior, formalizing it as a contextual bandit problem with retweeting as actions and follower changes as rewards. Results from a month-long experiment with 60 agents showed that aggregating experience across agents can reduce prediction accuracy and that Twitter's response to actions is non-stationary.
Twitter, a popular social network, presents great opportunities for on-line machine learning research. However, previous research has focused almost entirely on learning from passively collected data. We study the problem of learning to acquire followers through normative user behavior, as opposed to the mass following policies applied by many bots. We formalize the problem as a contextual bandit problem, in which we consider retweeting content to be the action chosen and each tweet (content) is accompanied by context. We design reward signals based on the change in followers. The result of our month long experiment with 60 agents suggests that (1) aggregating experience across agents can adversely impact prediction accuracy and (2) the Twitter community's response to different actions is non-stationary. Our findings suggest that actively learning on-line can provide deeper insights about how to attract followers than machine learning over passively collected data alone.