Set up initial sections in Quandary.
Began writing the code that will handle processing the interviews and generating analyses. So far, I've enhanced the HTML output to show word-counts next to each category of annotation; I need to update that to add a percentage of the content, although that might better be done as pre-processing for the XML. I've also done some file-name normalization (there was a lot of variation that got in the way of matching and decoding dates etc.), and I've also written a routine that extracts a plain-text version of all the utterances of the interviewee, so that they can be subject to similarity metrics and analyses of various kinds.
Also: team meeting today, with discussion of where we go with the analyses; went over a presentation with LG, and started planning an interactive questionnaire that we might present at DH 2019 along with some of our findings.
They came in with the print layout borked again, so I had to re-create them. Signed and submitted, Sept 1 and 2, but I'm waiting on Oct 1.
Remember that ${pd} is no good for Windows; you need ${pdu} wherever possible, otherwise ant tasks and transformations may fail.
I've finished a fairly straightforward interface that enables you to turn on and off highlighting for all the categories and subcategories in a text. It's pretty flexible and extensible, although at the moment the highlight colours for the base categories are hard-coded; making something self-adapting would actually make reconciliation between two encodings harder, because they'd end up with different highlighting. As it is, the base categories pretty much exhaust the easily-distinguished highlight colours. Waiting to see if EC can use this as it is, or whether she needs more features.
Met with EC to show her the documentation and the basic markup process. She'll now go away and pilot with DM.
Ahead of tomorrow's meeting with EC, completed a first pass through the documentation for the interview encoding, in the form of an ODD file, and added ODD processing to documentation to the build pipeline. Tweaked the CSS quite a bit, and then fed that back into the Keats project, where it originally came from.
Did a semi-manual conversion of the spreadsheet data into a first draft of the analysis taxonomy, and wrote an explanatory email (whose contents should end up in the documentation from the ODD file eventually) to DC and DM.
Recorded the final interview with CC, and processed the audio.
In between wrestling with vSphere and test-building vms, I've finished the first draft of my presentation on search engines for JADH. One more to go.