Improving business rating predictions using graph based features

Amit Tiroshi, Shlomo Berkovsky, Mohamed Ali Kaafar, David Vallet, Terence Chen, Tsvi Kufliky

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contributionpeer-review

16 Citations (Scopus)


Many types of recommender systems rely on a rich ensemble of user, item, and context features when generating recommendations for users. The features can be either manually engineered or automatically extracted from the available data, such that feature engineering becomes an important part of the recommendation process. In this work, we propose to leverage graph based representation of the data in order to generate and automatically populate features. We represent the standard user-item rating matrix and some domain metadata, as graph vertices and edges. Then, we apply a suite of graph theory and network analysis metrics to the graph based data representation, in order to populate features that augment the original user-item ratings data. The augmented data is fed into a classier that predicts unknown user ratings, which are used for the generation of recommendations. We evaluate the proposed methodology using the recently released Yelp business ratings dataset. Our results indicate that the automatically populated graph features facilitate more accurate and robust predictions, with respect to both the variability and sparsity of ratings.

Original languageEnglish
Title of host publicationIUI 2014
Subtitle of host publicationProceedings of the 19th International Conference on Intelligent User Interfaces
Place of PublicationNew York, NY
PublisherAssociation for Computing Machinery
Number of pages10
ISBN (Print)9781450321846
Publication statusPublished - 2014
Externally publishedYes
Event19th International Conference on Intelligent User Interfaces, IUI 2014 - Haifa, Israel
Duration: 24 Feb 201427 Feb 2014


Conference19th International Conference on Intelligent User Interfaces, IUI 2014


  • Feature Extraction
  • Graph-Based Recommendations
  • Recommender Systems


Dive into the research topics of 'Improving business rating predictions using graph based features'. Together they form a unique fingerprint.

Cite this