NDCG@K
PulseAugur coverage of NDCG@K — every cluster mentioning NDCG@K across labs, papers, and developer communities, ranked by signal.
-
Group recommendation evaluation bias uncovered, impacting reported progress
A new paper published on arXiv highlights a potential evaluation bias in group recommendation systems. The research demonstrates that standard metrics like HR@K and NDCG@K can be overly sensitive to how ties are resolve…
-
TimeRoute research tackles evolving modality relevance in recommendations
A new research paper introduces TimeRoute, a novel diffusion-based recommender system designed to address the challenge of changing modality relevance over time. TimeRoute employs a temporal-aware modal router to person…
-
TimeRoute system personalizes multi-modal recommendations by adapting to temporal shifts
Researchers have developed TimeRoute, a novel diffusion-based recommender system designed to address the challenge of time-varying modality usefulness in multi-modal recommendations. Unlike previous methods that use sta…
-
New PCA-GAT method enhances industrial process plan recommendations
Researchers have developed PCA-GAT, a novel approach for recommending machining process plans by integrating factual and normative industrial knowledge. This method treats process plan recommendation as a knowledge grap…
-
SalesLoop RL framework boosts sales lead ranking performance
Researchers have developed SalesLoop, a novel reinforcement learning framework designed to improve sales lead ranking by bridging the gap between offline model accuracy and real-world performance. The framework addresse…
-
Retrieval Evaluation Metrics Explained: P@K, MRR, NDCG
This article explains key retrieval evaluation metrics used to assess the performance of retrieval systems, including Precision@K (P@K), Recall@k, Mean Reciprocal Rank (MRR), and Normalized Discounted Cumulative Gain (N…