Thesis Sistemas de recomendación basados en métodos de filtrado colaborativo
Loading...
Date
2015-11
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Los Sistemas de Recomendación ayudan a la gente a tomar decisiones frente a grandes volúmenes de información. Una de las técnicas más populares es el Filtrado Colaborativo, el cual utiliza a usuarios con gustos similares para hacer recomendaciones. Se divide en dos categorías principales: algoritmos basados en memoria y basados en modelos. En este documento, se estudian, los métodos más representativos de cada categoría en el Estado del Arte. Luego, se presenta un estudio exhaustivo, a través de experimentos en tareas de predicción de ratings y recomendación de ítems, sobre diversos contextos. Finalmente, se obtienen conclusiones relevantes que pueden servir como base para entender y guiar investigaciones en esta área.
Recommender Systems help people to make decisions about large volumes of information. One of the most popular techniques is Collaborative Filtering, which uses users with similar preferences to make recommendations. It is divided into two main categories: memory-based and model-based. In this paper, we study the most representative methods of each category in the State of the Art. Then, a comprehensive study is presented, through experiments in prediction and recommendation tasks on various contexts. Finally, we obtain relevant conclusions that can serve as a basis for understanding and guiding research in this area.
Recommender Systems help people to make decisions about large volumes of information. One of the most popular techniques is Collaborative Filtering, which uses users with similar preferences to make recommendations. It is divided into two main categories: memory-based and model-based. In this paper, we study the most representative methods of each category in the State of the Art. Then, a comprehensive study is presented, through experiments in prediction and recommendation tasks on various contexts. Finally, we obtain relevant conclusions that can serve as a basis for understanding and guiding research in this area.
Description
Keywords
Filtrado colaborativo, Sistemas de recomendación, Álgebra lineal
Citation
Campus
Campus Santiago San Joaquín