RecSim: A Configurable Simulation Platform for Recommender Systems
This platform addresses the need for flexible simulation environments in recommender systems research, facilitating academic-industrial collaboration, though it is incremental as it builds on existing simulation concepts.
The authors introduced RecSim, a configurable simulation platform for recommender systems that supports sequential user interactions, enabling researchers to model various aspects of user behavior and item structure to advance reinforcement learning and recommendation techniques.
We propose RecSim, a configurable platform for authoring simulation environments for recommender systems (RSs) that naturally supports sequential interaction with users. RecSim allows the creation of new environments that reflect particular aspects of user behavior and item structure at a level of abstraction well-suited to pushing the limits of current reinforcement learning (RL) and RS techniques in sequential interactive recommendation problems. Environments can be easily configured that vary assumptions about: user preferences and item familiarity; user latent state and its dynamics; and choice models and other user response behavior. We outline how RecSim offers value to RL and RS researchers and practitioners, and how it can serve as a vehicle for academic-industrial collaboration.