Luke Roberts
Shena Hoffmann
Reference Information
Title: Improving meeting summarization by focusing on user needs: a task-oriented evaluation
Authors: Pei-Yun Hsueh, Johanna D. Moore
Presentation Venue: IUI 2009: Proceedings of the 14th international conference on Intelligent user interfaces; February 8-11, 2009; Sanibel Island, Florida, USA
Summary
This paper discusses a new approach to improving the summarization of meetings. The researchers discuss two types of summaries. The first provides a general summary of the meeting and is the type of summary produced by current systems (according to the paper, the general summary represents 30-40% of the meeting). The second provides a more decision-focused summary that is shorter than the general summary (according to the paper, the decision-focused summary represents 1-2% of the meeting).
The researchers performed a study in which they provided participants with four meetings through a Meeting Brower Interface. They asked the participants to summarize the decisions made in the meetings. Participants were randomly assigned one of four summary displays that were embedded into the browser interface that presented different information about the meeting. Some of the descriptions provided general summaries while others provided information generated by an algorithm described in another paper that pinpoints decision-related dialogue acts. In the study they focused on task effectiveness, report quality and user perceived success.
The researchers found that displaying decision-focused summaries were more effective and helped users get a better overview of the meeting. Even decision-focused summaries generated by the algorithm were more effective than general summaries created manually. However, it was found that decision-focused summaries written manually were still more effective than the ones generated by the algorithm.
Discussion
This sounds like it could be very useful for people who miss meetings or want to review meetings after they have occurred. As with most of the papers that I read in this class, I’m always left wishing I could know more about the actual algorithm they used, but the information provided in this paper was still interesting and proved that the algorithm and their interface can be useful for users.
As far as future studies, the researchers could focus on further improving the algorithm and they could run more experiments on meetings that are more and less structured to better identify the strengths and weaknesses in their current algorithm.
 |
| Image from paper of the Browser Interface |
It would definitely be interesting if the algorithm was more detailed to see what exactly they did. Also it seems like summarizers are always quite awkward so it would be nice to know how it picks out key points.
ReplyDeleteThis sounds like it could be helpful for recording and catching up with meetings. However, it sounds like it would be more useful to just have someone manually write a decision-focused summary rather than one generated by the algorithm, since that was more effective.
ReplyDeleteHaha, I didn't think about it being used for people to catch up on meetings that they missed. I suppose that could be good, but it may encourage people to skip meetings when they can just get the summaries of the decisions.
ReplyDelete