Welcome to the Folkopinion project
Folkopinion project hopes to discover
whether or not certain document and text classification methods can be used to classify (or group)
large number of blogs automatically without priori training data.
This is the research of one Doctoral student in Unitec New Zealand.
This research will ultimately be published as a thesis to complete his Doctoral degree.
Blogs are of interest in many-to-many communication studies because blogs provide a more user-friendly
and easier way to publish thoughts and opinions of individuals or groups. Unlike newsgroups however,
Blogs are not limited to just text-based content. Furthermore, websites such as YouTube even allows people to publish their opinions
in video-format (i.e. Video Blog).
Gaps in current research revolve around collection (and analysis) of large sample data from blogs, and to conduct research on differing context of existing methods.
Possible research areas include context-analysis of blogs to identify the authors,
addressing the shortcomings of current folksonomy systems, and content-analysis to query user
opinions from blogs.
Applications of Folkopinion include the ability to group "similar-voicing" opinions from blogs
for various business purposes (such as for a product or a politician), opinion-trend analysis of
blogs, and query on blogs to discover the effectiveness of a recent marketing campaign.
For further information please contact the researcher.