Active Learning Meets Optimized Item Selection

TL;DR

This paper introduces a combinatorial optimization framework that enhances recommendation system data collection efficiency by integrating optimized item selection with active learning, using advanced optimization and clustering techniques.

Contribution

It formulates a new optimization problem for item selection and develops a multi-level framework combining discrete optimization, clustering, and embeddings, advancing recommendation system experimentation.

Findings

Optimized item selection reduces data collection time.

Integration with active learning improves exploration efficiency.

Framework combines multiple advanced techniques effectively.

Abstract

Designing recommendation systems with limited or no available training data remains a challenge. To that end, a new combinatorial optimization problem is formulated to generate optimized item selection for experimentation with the goal to shorten the time for collecting randomized training data. We first present an overview of the optimized item selection problem and a multi-level optimization framework to solve it. The approach integrates techniques from discrete optimization, unsupervised clustering, and latent text embeddings. We then discuss how to incorporate optimized item selection with active learning as part of randomized exploration in an ongoing fashion.