A Bayesian Approach to Conversational Recommendation Systems
This work addresses the problem of improving recommendation accuracy and efficiency in conversational systems for users and platforms like stagend.com, but it appears incremental as it builds on existing Bayesian and information-theoretic methods.
The authors tackled the problem of conversational recommendation systems by developing a Bayesian approach that updates item probabilities based on user interactions, using information-theoretic criteria to optimize dialogue and termination decisions. In a case study on stagend.com, this method showed advantages in recommendation quality and efficiency, though specific numerical improvements were not detailed.
We present a conversational recommendation system based on a Bayesian approach. A probability mass function over the items is updated after any interaction with the user, with information-theoretic criteria optimally shaping the interaction and deciding when the conversation should be terminated and the most probable item consequently recommended. Dedicated elicitation techniques for the prior probabilities of the parameters modeling the interactions are derived from basic structural judgements. Such prior information can be combined with historical data to discriminate items with different recommendation histories. A case study based on the application of this approach to \emph{stagend.com}, an online platform for booking entertainers, is finally discussed together with an empirical analysis showing the advantages in terms of recommendation quality and efficiency.