|
Title:
|
APPLYING PROBABILISTIC TOPIC MODELS TO BLOG COMMUNITIES |
|
Author(s):
|
Alexandru Berlea |
|
ISBN:
|
978-972-8924-93-5 |
|
Editors:
|
Pedro Isaías, Bebo White and Miguel Baptista Nunes |
|
Year:
|
2009 |
|
Edition:
|
2 |
|
Keywords:
|
Community mining, topic detection applications, probabilistic models, blogs |
|
Type:
|
Short Paper |
|
First Page:
|
324 |
|
Last Page:
|
328 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
The popularity of Internet communities makes the content generated by their users increasingly interesting for various
purposes, including business ones. Analyzing the typically huge amount of community generated content calls for
automatic methods, a fundamental task hereby being to automatically detect the discussion topics. One very promising
approach thereto are probabilistic topic models (PTMs). In this paper we report our experiences with applying PTMs to
one popular Internet community. We discuss and interpret practical decisions to take for the application of PTMs and the
results obtained on our data. The particularities of PTMs in the context of typical blog collections are addressed. In
particular, we suggest how topic detection can benefit from typical blog features and also how the topic information can
be exploited for better supporting communities and for gaining relevant information from them. |
|
|
|
|
|
|
