As web-enhanced courses become more successful, they put considerable burdens on instructors and teaching assistants. We present our work on developing software tools to support instructors by automatic assessment of pedagogical discussions. We are developing prototype measures of discussion quality that rely on the quantity of discussion contributions. We are also developing techniques for assessing discussion contributions automatically by mining discussion text. Using information retrieval and natural language processing techniques, our tools learn to detect the conversation focus of threaded discussions, classify topics of discussions, and estimate technical depth of contributions. The results from these assessment tools provide basis for the development of scaffolding and question answering techniques for pedagogical discourse.

The authors of this paper present a user contribution measurement (UCM) model that has been validated against student discussion forums in two different domains; the field of psychology and computer science. Their developed model utilises modelling approaches that include speech act classification, rhetoric analysis and topic identification / classification (a Rocchio-style classifier was used to derive topic profile vectors through automatic ontology induction). I have been investigating the use of some of these approaches in improving the UCM model developed in my Honours thesis but I have yet to branch into that area of literature. Interestingly, the authors also provide a metric, coherence, which measures the variance of topic shifts within each thread.

Five parameters were used in the developed UCM and these include (no weightings of parameters were mentioned):

1. Number of messages
2. Length of all the messages
3. Number of messages elicited
4. Number of threads initiated
5. Number of different threads participated

• Set of relational dialogue rules
• Theory of speech acts & speech act classification
• Rhetorical structure theory & rhetoric analysis
• Sentence-level parsing of discourse (SPADE)
• Topic identification & classification
• Rocchio-style classifier
• Coherence