Choice of Metrics used in Collaborative Filtering and their Impact on Recommender Systems

Serradilla García, Francisco; Bobadilla Sancho, Jesus; Sánchez, J.L. y Martínez Murciano, Eduardo (2008). Choice of Metrics used in Collaborative Filtering and their Impact on Recommender Systems. En: "2nd IEEE International Conference on Digital Ecosystems and Technologies", 26/02/2008-29/02/2008, Phitsanulok, Thailand. ISBN 978-1-4244-1489-5. pp. 432-436. https://doi.org/10.1109/DEST.2008.4635147.

Descripción

Título: Choice of Metrics used in Collaborative Filtering and their Impact on Recommender Systems
Autor/es:
  • Serradilla García, Francisco
  • Bobadilla Sancho, Jesus
  • Sánchez, J.L.
  • Martínez Murciano, Eduardo
Tipo de Documento: Ponencia en Congreso o Jornada (Artículo)
Título del Evento: 2nd IEEE International Conference on Digital Ecosystems and Technologies
Fechas del Evento: 26/02/2008-29/02/2008
Lugar del Evento: Phitsanulok, Thailand
Título del Libro: Digital Ecosystems and Technologies, 2008. DEST 2008. 2nd IEEE International Conference on
Fecha: 30 Septiembre 2008
ISBN: 978-1-4244-1489-5
Materias:
Palabras Clave Informales: Collaborative filtering, metrics, recommender systems, cosine, correlation
Escuela: E.U. de Informática (UPM) [antigua denominación]
Departamento: Sistemas Inteligentes Aplicados [hasta 2014]
Licencias Creative Commons: Reconocimiento - Sin obra derivada - No comercial

Texto completo

[img]
Vista Previa
PDF (Document Portable Format) - Se necesita un visor de ficheros PDF, como GSview, Xpdf o Adobe Acrobat Reader
Descargar (178kB) | Vista Previa

Resumen

The capacity of recommender systems to make correct predictions is essentially determined by the quality and suitability of the collaborative filtering that implements them. The common memory-based metrics are Pearson correlation and cosine, however, their use is not always the most appropriate or sufficiently justified. In this paper, we analyze these two metrics together with the less common mean squared difference (MSD) to discover their advantages and drawbacks in very important aspects such as the impact when introducing different values of k-neighborhoods, minimization of the MAE error, capacity to carry out a sufficient number of predictions, percentage of correct and incorrect predictions and behavior when attempting to recommend the n-best items. The paper lists the results and practical conclusions that have been obtained after carrying out a comparative study of the metrics based on 135 experiments on the MovieLens database of 100,000 ratios.

Más información

ID de Registro: 3170
Identificador DC: http://oa.upm.es/3170/
Identificador OAI: oai:oa.upm.es:3170
Identificador DOI: 10.1109/DEST.2008.4635147
URL Oficial: http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=04635147
Depositado por: Memoria Investigacion
Depositado el: 31 May 2010 08:33
Ultima Modificación: 04 Mar 2015 16:31
  • Open Access
  • Open Access
  • Sherpa-Romeo
    Compruebe si la revista anglosajona en la que ha publicado un artículo permite también su publicación en abierto.
  • Dulcinea
    Compruebe si la revista española en la que ha publicado un artículo permite también su publicación en abierto.
  • Recolecta
  • e-ciencia
  • Observatorio I+D+i UPM
  • OpenCourseWare UPM