In this paper the authors build on prior literature to develop an adaptive and time-varying metadata-enabled dynamic topic model (mDTM) and apply it to a large Weibo dataset using an online Gibbs sampler for parameter estimation. Their approach simultaneously captures the maximum number of...