Lightning talks from the first day of Code4libBC.
Provincial Digital Library Update (Daniel Sifton/Caroline Daniels)
Looking at what others have done
* DPLA: gave presentation at code4lib portland about old system. New system based on PostgreSQL, Solr, Ruby-driven…
* testing with VM using vagrant, ansible, elasticsearch, solr, ruby, rail console
* Supplejack: 2nd generation platform, Harvest OAI/etc., Apache-based manager, Working, API connected to Solr and MongoDB
* more during breakout
Building Quick and Dirty Library Tools with Google Apps Script (Terry Brady)
- Google Apps Script: server-side JS, access to Google’s API, run as Auth’ed user
- can package as Chrome extension/plugin
- available to where users are, sharable via Google Drive, script has configurable authorization (run as user vs. author)
- CVS files can be corrupted when shared between users, especially with autocorrect in excel/google spreadsheet
- basic web service where can upload CVS, make sure auto-correction is turned off, and uploaded to google drive
- can do lots of things especially using triggers, such as time or event based triggers
- can package as plugin into Chrome store
Makerspaces and the Maker Community in the Lower Mainland (Maryann Kempthorne)
- Vancouver Maker Foundation organizes Mini MakerFaire at PNE in June
- Surrey Public Library: Maker Meetups
- want to make sure libraries are involved in the movement
- VanMakerEd: meets 4 times a year; VPL usually hosts one per year
- 3D Maker Meetup at Richmond PL
- MakerSpace Fraser Valley joint programming with FVRL
- Vancouver HackSpace
- MakerLab (Strathcona)
- training for teachers and others
- Island: makerspace at Camosun, Greater Victorial PL, Nanaimo, Gibsons
- my addition: ladies learning code for teachers
Web Analytics with Piwik (Jonathan Schatz)
Slides
* alternative web analytics (instead of Google Analytics)
* free, self-host, LAMP stack, FOSS, modules, privacy focused
* data control, friendly defaults with sensitive private, honours donottrack
* minimizing data collected: anonymized IPs, set retention periods, offer opt-out, TLS
* standard metrics: visits, pageviews, bounce rate, visit & action, scopes, custom dimentions for reports and segments
* simple custom variable tally: example: keep track of which organization downloaded files from repository
* transition: visualize where traffic is coming from and where it goes out
* calls can be done using client-side, server-side; similar to GA
* goals/conversions: example: downloaded more than 3 items, watched a tutorial, referred by email; will then display metric x goal
* data quality: piwik spamlist updated by community and can be customized
* overall: strong privacy, mix of metrics, some config (can chock on high traffic, so might need to throttle number of writes), graphics aren’t great, but information can be exported
* good API: can build on top to pull reports and widgitized in other apps
* SFU doing some work on using API to bring in analytics into Islandora
Break time
TechBC Memory Project metadata upgrade (Holly Hendrigan)
- interviews 30 people that were part of the TechBC university (that no longer existed)
- uploaded audio files to Summit (IR) with minimal metadata
- oral history conundrum: transcription expensive, meaning lost when transcribed to text, but without transcription, will unlikely be used
- OHMS: oral history metadata synchronizer: interview summarized, segments can be summarized and given keywords
- SFU developer refractored viewer to work in Islandora
- site has basic browsersable pages, then metadata searchable, viewer has audio viewer, metadata tab, and transcript in PDF format
- lessons learned: too big a job for a single person to do
- oral history now accessible
Making Data Human: Visualizations using the Open Collections API (Schuyler Lindberg)
- goal: create visualizations to understand the data
- based on aggregation queries in ElasticSearch
- have pretty much full access to EleasticSearch functions with UBC OpenCollections API
- can neat aggregations to do data over time, and other cool things, assisted by use of filters
- also used for faceting in the UI such as list and count of subjects
- JSON is now necessarily optimal data format, but can be converted to tabular data
- can create visualizations using tools such as d3
- scripting language built-in so can do very refined comparisons
- used internally to identify bad metadata, example: 2 ‘still image’
Getting Things Done (Cynthia Ng)
I wrote up my presentation as usual, so please go see the blog post with the full slides and script.
Lunch Time
First lunch, then breakout sessions time!
Learning (Lib)Tech 