Slides at http://conferences.infotoday.com/documents/293/C205_Affelt.pptx
Data scientist #1 job at GlassDoor, but librarians can work with big data, too.
Gardner Vs: volume, velocity, variety, verification, value.
Examples:
* Server log files
* Social media content
* Digital images (security cameras, license plates)
* Geolocation from smartphones
* Internet of Things (sensors everywhere)
* Personal security data (e.g., Nest)
* Video (FB says it will be all video in 5 years; where will the transcripts and metadata come from?)
* Data that “used to be dropped on the floor”
What are we giving up for convenience?
Before, you would sample data; now you can just analyze all of it.
Consumer Reports is working on a consumer-protection standard for IoT products.
IoT considerations:
* Safety and reliability of devices
* Security ramifications
* Preservation of data
* Privacy issues
* Manufacturer obligation to export data?
Could people hack into your pacemaker or diabetic pump?
What about ransomware?
What could be considered evidence in court?
A.I. In legal:
* Cost estimates
* Document drafting
* E-discovery: Elevate, Logikcull
* Judicial characterization: How a judge might rule
* Contract review
Problematic data:
* Polls in 2016 predicting Clinton would win: problems with polling, same-day voting registration (polls use “registered voters”), caller ID (not a random sample)
* Brexit: turnout was incorrectly predicted
Roles for librarians:
* Visulization and report creation: Tableau OPublic, Inforgr.am, Adobe Creative Cloud (Inkscape is a free alternative), librariandesignshare.org
* Research verification
* New class of expert
* Alogrithm accountability
* Municipal government/community organization liaison
* Alerts: news on big data for your org.
* Training in big data tools