Reference information
Authors: Robert Speer, Catherine Havasi, Nichole Treadway, Henry Lieberman
When/Where: IUI’10, February 7–10, 2010, Hong Kong, China.
Summary
This paper introduced Luminoso, a tool for helping researchers visualize and understand a semantic space, like in forum posts or survey responses, using natural language processing. The objective is to help analyze large datasets to understand it. The system can open sets of documents, some of which can be tagged as canonical to determine whether other documents agree with it semantically. This is done with natural language processing to find patterns and form conclusions from the documents.
Documents are represented in a 2D space, where each part of a document corresponds to a point in the space. The size of each point indicates the number of times that the item appears in the input and has a text label that describes a common feature or name of a document. Users can select, or "grab", points to change the projection of the data and learn more about that point. The colors of the points indicate correlation to the grabbed point.
Discussion
I thought Luminoso was very useful and could have many applications in the future for visualizing data to find the right connections between them. I would have liked to see a user study conducted to see how well the system works, but I think the idea and concept is important by itself. This paper reminded me somewhat of Opinion Space, a web application developed by Dr. Ken Goldberg, who gave a seminar lecture earlier this semester. The interface and fundamental design from both Luminoso and Opinion Space are similar in that they attempt to relate large datasets, like documents and opinions or comments, in a 2D space. The paper, particularly the conclusion, reminded me of Chapter 7 Skim in "Why We Make Mistakes".
No comments:
Post a Comment