A manifesto for modeling and measurement in social media

Graham Cormode, Balachander Krishnamurthy, Walter Willinger


Online Social Networks (OSNs) have been the subject of a great deal of study in recent years. The majority of this study has used simple models, such as node-and-edge graphs, to describe the data. In this paper, we argue that such models, which necessarily limit the structures that can be described and omit temporal information, are insufficient to describe and study OSNs. Instead, we propose that a richer class of Entity Interaction Network models should be adopted. We outline a checklist of features that can help build such a model, and apply it to three popular networks (Twitter, Facebook and YouTube) to highlight important features. We also discuss important considerations for the collection, validation and sharing of OSN data.


online social networks; modeling; measurement

Full Text:


DOI: https://doi.org/10.5210/fm.v15i9.3072

A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2019. ISSN 1396-0466.