04 November 2010

Discovering Genres of Online Discussion Threads via Text Mining

This article caught my eye after last night's class on voice threads. It was also educational for me to find because I used some of the data base search techniques we learned in the other class. Anyway the link is: Discovering Genres... This article was about teachers using things like discussion boards and maintaining them. The article points out how much work goes in to monitoring the discussion boards, maintaining them, looking at students work as well as trying to participate and be available for questions. The article's research question was whether or not a certain web mining program can help with the process of data collection and is it as accurate as a human judge.


The first thing the program would do is place posts into different genres, such as announcements, questioning, clarification, interpretation, conflict, and assertion therefore making it easier for the teacher to go through the information. Moodle is a program we learned about in class that educators use for these types of forums. The great thing is, the students have access to supplementary materials, can work from a place other than school and can ask questions.  Of course the teacher is the one that has to do the coding and reading of every post which the article argues is very time consuming and is usually done on personal time.


The article is really detailed and computer science sounding. It was hard to read about the data collection because it got pretty technical. It talked about how different fonts or language fonts could throw off coding results. I recommend this article to more technical people who want to know how things work. I just want to know what works and what is fun to use.

No comments:

Post a Comment