Portrait de Andres Ferraro n'est pas disponible

Andres Ferraro

Alumni

Publications

Commonality in Recommender Systems: Evaluating Recommender Systems to Enhance Cultural Citizenship
Andres Ferraro
Gustavo Ferreira
Fernando Diaz
Georgina Born
2023-02-22
ArXiv (prépublication)
doi.org
arxiv.org
Measuring Commonality in Recommendation of Cultural Content: Recommender Systems to Enhance Cultural Citizenship
Andres Ferraro
Gustavo Ferreira
Fernando Diaz
Georgina Born
2022-09-13
Proceedings of the 16th ACM Conference on Recommender Systems (publié)
doi.org
arxiv.org
Offline Retrieval Evaluation Without Evaluation Metrics
Fernando Diaz
Andres Ferraro
Offline evaluation of information retrieval and recommendation has traditionally focused on distilling the quality of a ranking into a scala… (voir plus)r metric such as average precision or normalized discounted cumulative gain. We can use this metric to compare the performance of multiple systems for the same request. Although evaluation metrics provide a convenient summary of system performance, they also collapse subtle differences across users into a single number and can carry assumptions about user behavior and utility not supported across retrieval scenarios. We propose recall-paired preference (RPP), a metric-free evaluation method based on directly computing a preference between ranked lists. RPP simulates multiple user subpopulations per query and compares systems across these pseudo-populations. Our results across multiple search and recommendation tasks demonstrate that RPP substantially improves discriminative power while correlating well with existing metrics and being equally robust to incomplete data.
2022-07-07
Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval (publié)
doi.org
arxiv.org
Offline Retrieval Evaluation Without Evaluation Metrics
Fernando Diaz
Andres Ferraro
Offline evaluation of information retrieval and recommendation has traditionally focused on distilling the quality of a ranking into a scala… (voir plus)r metric such as average precision or normalized discounted cumulative gain. We can use this metric to compare the performance of multiple systems for the same request. Although evaluation metrics provide a convenient summary of system performance, they also collapse subtle differences across users into a single number and can carry assumptions about user behavior and utility not supported across retrieval scenarios. We propose recall-paired preference (RPP), a metric-free evaluation method based on directly computing a preference between ranked lists. RPP simulates multiple user subpopulations per query and compares systems across these pseudo-populations. Our results across multiple search and recommendation tasks demonstrate that RPP substantially improves discriminative power while correlating well with existing metrics and being equally robust to incomplete data.
2022-04-25
ArXiv (preprint)
doi.org
arxiv.org