Linguistically Guided Community Discovery

What is it about?

This paper uses community discovery algorithms on graphs of texts collected from online topic communities to discover factions, or sub-communities which frequently take the same side on an issue. The graphs used capture similarity relations among the texts produced by members of the online communities. The similarity graphs are produced by summing similarity graphs of different kinds, in particular similarities of language use and similarities of websites linked to.

Featured Image

Why is it important?

This paper provides an automatic tool for discovering factions in topic communities.