Fair and Balanced: Learning to Present News Stories

Relevance, diversity and personalization are key issues when presenting content which is apt to pique a user's interest. This is particularly true when presenting an engaging set of news stories. In this paper we propose an ecient algo- rithm for selecting a small subset of relevant articles from a streaming news corpus. It oers three key pieces of improve- ment over past work: 1) It is based on a detailed model of a user's viewing behavior which does not require explicit feed- back. 2) We use the notion of submodularity to estimate the propensity of interacting with content. This improves over the classical context independent relevance ranking al- gorithms. Unlike existing methods, we learn the submodu- lar function from the data. 3) We present an ecient online algorithm which can be adapted for personalization, story adaptation, and factorization models. Experiments show that our system yields a signicant improvement over a re- trieval system deployed in production.