Community detection over feature-rich information networks: An eHealth case study

In this paper, we present a novel graph data model to analyze eating habits and physical activities of a large number of persons, aiming at automatically detect groups of users sharing the same lifestyle using Social Network Analysis facilities. We focus our attention on physical activities and dietary habits of users because they often can be correlated to several types of diseases. Indeed, they constitute a real example of feature-rich information network (containing multi-relational and heterogeneous data) that can support different analytics. Furthermore, a novel community detection approach has been exploited to detect groups of users sharing same behaviors/habits within the obtained information network by leveraging nodes' and edges' properties. Finally, an extensive experimentation on simulated and real networks has been performed for evaluating the proposed approach in terms of efficiency and effectiveness, outperforming some of the most diffused state-of-the-art approaches (up to 8%). (c) 2022 Elsevier Ltd. All rights reserved.

Automated summary

This paper presents a novel graph data model for analyzing eating habits and physical activities of individuals, with the aim of automatically detecting groups of users who share the same lifestyle using Social Network Analysis. The study focuses on physical activities and dietary habits as they often have correlations with various diseases.