We identify shortcomings of current recommender systems.
We present an interactive recommender framework to tackle the shortcomings.
We analyze existing interactive recommenders along the dimensions of our framework.
Based on the analysis, we identify future research challenges and opportunities.