I think concordances and visualizations would be great additions to our catalogs and "discovery systems".
In my personal "sandbox" I have implemented concordances against just about every item in my Alex Catalogue -- http://bit.ly/hF5oy8 Display the most common n words in a text. For any given letter, display all the words starting with that letter and count the number of times they occur. Display and count the most frequent n-grams. Search for any word (or regular expression) in a text. Display the results surrounded by the words in context. All of this functionality enables the reader to see overall patterns in the text. I have also begun to incorporate these ideas in an implementation of VuFind -- http://bit.ly/aeIUqV
Us librarians love words, but as the saying goes, pictures are worth thousands of words. Consequently I have also begun to visualize concordances using "network diagrams". Nodes in the illustration represent words. Clusters of nodes represent proximity. Lines in the illustration represent relationships. The whole thing is analogous to visualizing what a person says "in the same breath" as a given word -- http://bit.ly/gD12og Unfortunately, the visualization does not work on Internet Explorer.
Concordances are really old implementations of text mining. Visualizations convey ideas quickly. There is an increasing abundance of full text materials in our collections. Combining all of these things together in our "discovery systems" will enhance their functionality. This is an opportunity for libraries to shine. Fun!?
--
Eric Lease Morgan
Hesburgh Libraries, University of Notre Dame
Great Books Survey -- http://bit.ly/auPD9Q
Received on Fri Jan 14 2011 - 08:53:16 EST