6th Place Solution to Amazon KDD Cup ’23 Task2: Next Product Recommendation for Underrepresented Languages

Chiaki Ichimura, Daiki Toda

Keywords: Recommender Systems, Collaborative Filtering, Content Based Filtering, Learning to Rank, KDD Cup

Abstract: In this paper, we present our solution, which achieved 6th place in the Amazon KDD Cup Task2 (Next Product Recommendation for Underrepresented Languages) in 2023. Our recommendation pipeline comprises two stages: candidate generation through several recommendation algorithms and ranking using machine learning models. We incorporate predictions from multiple individual recommender models, combining them based on prediction ranks, and achieve a Mean Reciprocal Rank (MRR) of 0.44798 on the competition leaderboard.

Submission Number: 9