Content-Based Recommendation for Stacked-Graph Navigation

Abstract

AbstractIn this chapter we present a system which combines interactive visual analysis and recommender systems to support insight generation for the user. Collaborative filtering is a common technique for making recommendations; however, most collaborative-filtering systems require explicit user ratings and a large amount of existing data on each user to make accurate recommendations. In addition, these systems often rely on predicting what users will like based on their similarity to other users. Our approach is based on a content-based recommender algorithm, where promising stacked-graph views can be revealed to the user for further analysis. By exploiting both the current user navigational data and view properties, the system allows the user to see unseen-views suggested by our system. After testing with more than 30 users, we analyze the results and show that accurate user profiles can be generated based on user behavior and view property data.KeywordsDwell TimeTextual AttributeUser ProfileNumerical AttributeVisualization SystemThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.