Home

This is intended to be an internal set of documents to help organize the livejournal text analysis project.

There's a discussion group called berkeley-dsc under http://msdse.slack.com (you can install a local app, or just use the browser).

Resources

1. John's software:

• BID (Berkeley Institute of Design) Data project: http://bid2.berkeley.edu/bid-data-project/
• The tooling itself: https://github.com/BIDData/BIDMach (parsing code in flex is in ./src/main/C/newparse)
• On stout, look in /big/code/BIDMach/bin for the parsing code
• Currently installed under /opt on mercury (and parsing code is also under BIDMach.../bin, CUDA is working.

1. Info on the research done so far on Stress Management (it does not have a lot on data mining): http://bid.berkeley.edu/stressmanagement/

2. Life Events Questionnaire (LEQ), which is used to evaluate chronic stress:

• Survey: http://nursing.ucsf.edu/sites/nursing.ucsf.edu/files/LifeEventsQues.pdf
• Pablo's document with potential basis for the regular expressions that will represent LEQ (the examples here come from sampling over the data manually, so it is not exhaustive at all) is in our "Destress" google drive folder.

1. See here for location of the full data set. Sample raw data form the livejournal database is also in our google drive folder (only one directory out of 1282 directories with about 100 to 200 users per subdirectory).