2014 | OriginalPaper | Buchkapitel
Testing Community Detection Algorithms: A Closer Look at Datasets
verfasst von : Ahmed Ibrahem Hafez, Aboul Ella Hassanien, Aly A. Fahmy
Erschienen in: Social Networking
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Social networks of various kinds demonstrate a strong community effect. Actors in a network tend to form closely-knit groups; those groups are also called communities or clusters. Detecting such groups in a social network (i.e., community detection) remains a core problem in social network analysis. Among the challenges that face the researchers to come up with advanced community detection methods, there is a key challenge, which is the validation and evaluation of their methods. The limited benchmark data available, the lack of ground truth for many of the available network datasets, and the nature of the social behavior factor in the problem, turned the evaluation process to be very hard. Accordingly, understanding such challenges may help in designing good community detection methods. This chapter presents testing strategies for community detection approaches and explores a number of datasets that could be used in the testing process as well as stating some characteristics of those datasets.