Learning to Rank in the Position Based Model with Bandit Feedback
This work addresses bias in personalization for online content ranking, but it is incremental as it builds on existing bandit algorithms.
The paper tackled the problem of bias in supervised learning-to-rank methods for content ranking by proposing novel extensions of LinUCB and Linear Thompson Sampling algorithms using contextual multi-armed bandits and the position-based click model, and validated them through offline experiments on synthetic datasets and online A/B tests.
Personalization is a crucial aspect of many online experiences. In particular, content ranking is often a key component in delivering sophisticated personalization results. Commonly, supervised learning-to-rank methods are applied, which suffer from bias introduced during data collection by production systems in charge of producing the ranking. To compensate for this problem, we leverage contextual multi-armed bandits. We propose novel extensions of two well-known algorithms viz. LinUCB and Linear Thompson Sampling to the ranking use-case. To account for the biases in a production environment, we employ the position-based click model. Finally, we show the validity of the proposed algorithms by conducting extensive offline experiments on synthetic datasets as well as customer facing online A/B experiments.