Visualizing co-occurances with force-directed diagrams The contents of this directory demonstrate how co-occurances in a text can be visualized through force-directed diagrams. Concordances are a great way to locate a word in a text and view the words on either side of it. They are literally Medieval keyword-in-content (KWIC) indexes. The contents of this directory builds on the functionality of concordances by identifying the words around a given keyword, counts their frequency, builds a matrix from them, converts the matrix into a JSON data structure, and displays it as a force-directed diagram. The result is an implmentation of "distant reading" as coined by Franco Moretti in Graphs, Maps, Trees: Abstract Models for Literary History. The scripts create graphics representing concepts in the whole or part of a text. Using this tool a person can do finer grained "close reading" by previewing and reviewing concepts before and after the traditional reading process. Considering the increasing amount of freely available full text on the Internet, and considering the ubiquitous availability of computing horse-power, implementations of distant reading are natural conequences. Contents The contents of this directory include: * readme.txt - this file * bin - the executable CGI scripts * etc - the texts used for analysis * lib - the Javascript used to render the graphic * src - the CGI scripts in a browser-viewable form Acknowledgements Thanks go to Michael Clark who showed me my first network diagrams. Thanks goes to Ed Summers who shared a few of his force-directed diagrams with the Code4Lib Community. Thanks go to the folks at the Stanford Visualization Lab who created Protovis. -- Eric Lease Morgan December 28, 2010