VisualRS: Java framework for visualization of recommender systems information
作者:
Highlights:
•
摘要
VisualRS is an object oriented Java framework that allows: 1) To relate items or users from recommender systems, 2) To provide tree graph structures containing the most relevant relationships between items or between users, 3) To create visualizations, using Gephi, of the resulting tree graphs, 4) To select the tree graph structures that allow a more balanced visualization and an easier navigation, 5) To choose between an assortment of collaborative filtering similarity measures, 6) To measure the quality of the results making use of centrality measures in graphs, and 7) To easily embed new similarity metrics and quality measures.VisualRS can be used: 1) As recommender systems analysis tool, 2) To provide users and technical managers with a visualization tool of items and users, and 3) To facilitate research both into the collaborative filtering graphic representation and into the information navigation fields. It provides a broad basis for personalization and experimentation.
论文关键词:Related information,Tree graphs,Similarity measures,Collaborative filtering,Recommender systems
论文评审过程:Received 28 September 2017, Revised 23 April 2018, Accepted 24 April 2018, Available online 30 April 2018, Version of Record 28 May 2018.
论文官网地址:https://doi.org/10.1016/j.knosys.2018.04.028