Improving Online Help Fora by identifying Frequently Asked Questions and Answers

Most universities and colleges employ web-based Learning Management Systems, such as Blackboard or Canvas, to promote discussion amongst students enrolled in their undergraduate units. Over a number of years, the archives of most units' discussions are retained, often in a standard format, such as the "mailbox format" described in RFC2822 (http://www.faqs.org/rfcs/rfc2822.html). This is the format used by our own cshelp software, which has has retained units' discussions for nearly twenty years.

Naturally, a large number of Frequently Asked Questions are asked by each year's student cohort and unit coordinators answer these each year "from memory", or with reference to the previous years' archives to provide a consistent response across years. Even within a single semester, forum discussions in large units become overrun with repeated questions. Students who do not read their units' fora every day find it difficult to keep up with the volume of discussion, and simply ask the same questions, again, rather than searching for an answer to their problem.

This project will investigate if online help fora can be improved by:

(1) investigating if Frequently Asked Questions, and their answers, can be automatically identified within discussion archives, and then offered as possible answers to the questions from new students each year, and

(2) checking students' new forum questions, to determine if their question is strongly related to any existing discussions, in an attempt to reduce the incidence of students actually asking Frequently Asked Questions.

The project will likely employ Natural Language summarisation and matching modules (already written) in Python. The project must employ an evaluation component, where current CSSE students are invited to comment on the effectiveness of the identified answers.

Client


Contact: Chris McDonald
Phone: 6488 2533
Email: [email protected]
Preferred contact: Email
Location: CSSE Rm 2.20

IP Exploitation Model


The IP exploitation model requested by the Client is: Creative Commons (open source) http://creativecommons.org.au/


Department of Computer Science & Software Engineering
The University of Western Australia
Last modified: 22 July 2021
Modified By: Michael Wise 		UWA