Researchers have developed Decision4Rec, a novel framework designed to integrate decision models into recommender systems. This framework utilizes knowledge graphs and active graphs to log recommendation scenarios and employs a pool of agents to propose candidates, with a residual ranker selecting the top options. Decision models, such as the commercial Jev, are trained on real choices to provide probabilities for each option, which are then combined with the ranker's scores. The framework has demonstrated significant improvements in recommendation accuracy, outperforming baselines by up to 36% in NDCG@10 on sparse datasets and nearly matching baselines on dense datasets. AI
IMPACT This framework could enhance the accuracy and efficiency of recommender systems by leveraging decision models and knowledge graphs.
RANK_REASON The cluster describes a new research framework and its performance on recommendation tasks. [lever_c_demoted from research: ic=1 ai=1.0]
Read on arXiv cs.IR (Information Retrieval) →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
