I have been provided additional prompts from Dr. Jenkins.

• What have you accomplished?

So far, I have researched potential software to aid in cataloging and concordancing the selected range of texts in order. This has included software such as AntConc, Corpkit, and WordSmith. Antconc was selected on the basis that the software has multiplatform support and ease of use; unfortunately, AntConc does not support annotation; however, such annotation’s accuracy would be very poor/inaccurate on the earlier scans due to their poor quality and the inadequacies of Adobe’s OCR technology. I had petitioned with Amazon’s Web Services to be included in the closed ‘beta’  of their Amazon Textract which is documents to automatically extract text and data from scanned documents. I never heard back. I researched the selected excerpts of Dr.Stone and additional reading of the target documents to build a list of potential context keywords which would build a concordance relevant to the project – This was straight forward enough: use Adobe’s OCR technology to then export the PDFs to TXT file which would then be used by the concordance. This would use a basic form of Regex in order to build a list that would find the lemma word of selected words and also fetch any of the same with additional affixing as well as tense (eg. Teach/teaching, taught,). The original list returned a plethora of false positives and as such required a prioritization of search terms which fell into the following (See next figure); additionally, Adobe’s OCR technology was rather inaccurate in the earlier scanned documents and I continually had to return to the PDF, or exported TXT file, to make a correction.

Once built, the concordance was exported to Excel and item by item I cross-referenced the PDF to the central concordance list and created annotated highlights within the original document. This was very time-consuming.

A naming convention was developed to easily identify the context of the excerpts: The document’s commission acronym, Year of Publication, Perspective or role in the project, location of the report. E.g. ARBIC-1872-P-Haines-[Author] In this example ARBIC stands for the Annual Report to the Board of Indian affairs Commission and P stands for policy.

• What have you learned?

Perfectionism absorbs an exponential amount of time and ultimately leads to low productivity and burnout when dealing with low-quality OCR’d PDFs.

• Are you meeting the goals for your internship goals?

The current explicit goal of researching, creating, and expanding the current corpus have been met. I will need to iron out the process and comment on how it was approached.

• Do you need to make any changes to your work habits? 

My work habits need to be far more consistent rather than these burst of productivity – it leads to a much larger mental burnout; additionally, some approaches taken when building the corpus and attempting to build a central document to access the information was short-sighted and I find myself having to go back and redesign the approach in order to facilitate ease of user access later one after I have finished the internship as I would want to fruitless efforts or creating an end item which is a hassle to grasp.

Included is a table with the first revision of context keywords used to create the excerpt report. ‘?’ signifies a wildcard for any one character. ‘+’ denotes a wildcard for zero or more characters.