Comments
http://shennessy11.blogspot.com/2011/03/paper-reading-16.html
http://stuartjchi.blogspot.com/2011/03/paper-reading-16-creating-collections.html
Reference Information
http://shennessy11.blogspot.com/2011/03/paper-reading-16.html
http://stuartjchi.blogspot.com/2011/03/paper-reading-16-creating-collections.html
Reference Information
Title: Finding Your Way in a Multi-dimensional Semantic Space with Luminoso
Author: Roobert Speer, Catherine Havasi, Nicole Treadway, Henry Lieberman
Publisher: IUI '10, February 7-10, 2010 Hong Kong
Summary
Language and understanding play a large role in HCI. There are many times when it may be helpful to visualize and understand the meaning of large collections of data in many different dimensions. Users express their opinions en masse on surveys, forums and dialogue systems. Computational linguistics is a way to solve the problem of understanding this sort of feedback and can often yield insights that normal statistics miss.
The researchers propose a system, Luminoso, to help users create and develop specialized semantic networks. Luminoso is an interactive application that aids researcher in exploring semantic spaces in an intuitive way intended to dimensionally reduce the available data. It creates a vector space from a folder of input documents and mines the data in an interactive way. It uses the power of LSA which relies on the co-occurence of words in the input documents, making semantic connections between words.
Discussion
In understand the idea behind this research but I had a lot of trouble reading this article. I felt like there were a lot of technical terms and such that I didn't understand. I also didn't really understand how the user interacted with the system. They mentioned a "grabbing" procedure to move the projected data points on the screen, but I wasn't sure how that dynamically worked across dimensions.
The researchers propose a system, Luminoso, to help users create and develop specialized semantic networks. Luminoso is an interactive application that aids researcher in exploring semantic spaces in an intuitive way intended to dimensionally reduce the available data. It creates a vector space from a folder of input documents and mines the data in an interactive way. It uses the power of LSA which relies on the co-occurence of words in the input documents, making semantic connections between words.
Discussion
In understand the idea behind this research but I had a lot of trouble reading this article. I felt like there were a lot of technical terms and such that I didn't understand. I also didn't really understand how the user interacted with the system. They mentioned a "grabbing" procedure to move the projected data points on the screen, but I wasn't sure how that dynamically worked across dimensions.
This sounds pretty confusing, but I do like the idea of using an interactive application to present data in a way that is easier to see and understand. It seems like it would be easier to analyze the data if it were presented visually.
ReplyDeleteThis article seems a little out there. too much technical stuff is always annoying, I agree
ReplyDeleteI also found this set of articles fairly technical. This paper seemed pretty interesting however, even if you're only looking at it from a overall generalization of it.
ReplyDeleteThis idea sounds great. I hope this is something big data can help with given that it requires a good deal of computation power.
ReplyDelete