Poincare: Recommending Publication Venues via Treatment Effect Estimation

Choosing a publication venue for an academic paper is crucial step in the research process. However, many cases, decisions are based solely on experience of researchers, which often leads to suboptimal results. Although there exist recommender systems papers, they recommend venues where expected be published. In this study, we aim from different perspective. We estimate number citations will receive if published each and has most potential impact. two challenges task. First, only one venue, thus, cannot observe would were another venue. Secondly, contents not statistically independent; that is, selection biases choosing venues. paper, formulate recommendation problem as treatment effect estimation problem. use bias correction method impact effectively papers highlight effectiveness our using data computer science conferences.