Voyer mining referenced in this post: http://voyant-tools.org/tool/Cirrus/?corpus=1397698027170.9588&query=&stopList=stop.en.taporware.txt&docIndex=0&docId=d1397620924900.b1f86013-5db2-1c47-b75c-9bf1ab5be0c2
This week, we focused on the text that is available online on the internet. Because there is such a wealth of Civil War documents available to us, it is beneficial to use word-mining to find the frequencies of significant words in the hundreds of thousands of words per historic document. I used Voyant to word-mine the following document, highlighted in blue, from Cornell Library’s catalogue. After omitting the display of stop words, the most common words were quite applicable to the war; general, army, cavalry, va, etc. When I searched the document for the frequency of words like “death,” “disease,” “casualty/ies,” and “injured,” their frequency count was much lower than I expected. “Death” appeared only 26 times in the 369,000+ word document, and casualty only appeared one time. This was surprising since I expected a document entitled “report” to include a count of casualties from battles.
In terms of the relation between these documents and my own soldier, Wilhelm Kurz. I searched for any indication of his 29th NY Infantry being mentioned and it never was. Through numerous keyword searches, I have been unable to relate these documents to my soldier. I am worried, now, about how I will incorporate the text-mining element, wether through Voyant on Google NGram, into my final Omeka exhibit. My plan is to continue mining this document and search for locations that my soldier was in during that time frame, hopefully finding reports on the activities in the area at that time. This, at the very least, will link my soldier’s records to those from the Cornell Library. I am interested
- Series I – Volume XIX – in Two Parts Chapter XXXI – Operations in Northern Virginia, West Virginia, Maryland, and Pennsylvania. Sep 3-Nov 14, 1862.
- Part II – Reports, September 20-November 14; Correspondence, etc., Sept 3-Nov. 14