Access 2014 Day 2: Afternoon Notes

We continue with the afternoon of day 2 of Access 2014.

On the program is linked data and some lightning talks.

Linked Data is People: Using Linked Data to Reshape the Library Staff Directory

  • Jason A. Clark, Head, Library Informatics & Computing, Montana State University
  • Scott W.H. Young, Digital Initiatives Librarian, Montana State University

Linkded Data Slides

Use case e.g. staff directory.

Linked Open Data: practice of publishing structured, machine-actionable data with an open license for sharing and reuse. (See some of @ahitchens ‘ intros to linked data).

goals:

  • create distributed flexible data systems
  • add context and dimension to data
  • demonstrate relationships based on standardized and reusable vocabulary

Great for mapping persons relationships

Measured by Five-star LOD scaled. Wanted to get to at least #3. Actually got to 5.

Project

  • applying concepts of linked data for our local library community
  • finding a real world use case for graph modeling using web ready open source tools
  • top considerations: speed and ease of use
  • how can our people be better discovered and understood?

RDF, SQL, PHP, D3.js

Goals: new data model, new UI, network visualization (showing connections between research interests and subject liaison), basic reomcmendation app, SEO results, 5-star LOD

Linked Data: RDF, OWL, SPARQL, SKOS, DBO, schema.org – which ones to use, how to use them? where are our people? and where are our concepts?

VIAF is not viable, but ORDCID in there for people. LCSH doesn’t have many of the concepts, so looked at dbpedia to describe concepts.

Ontology Model

  • Person Name – schema.org -> Person ID
  • Google Scholar ID / ORCiD -> Person ID
  • Person ID — Topic ID
  • Topic Name – schema.org -> Topic ID
  • dbpedia URI – sameAs -> Topic ID

the harder part was learning how people wanted to be described.

Why schema.org

  • libraries as publishers in the Web of Data
  • others were using it e.g. OCLC, wanted to be recognized
  • machine understanding of our data = relevance, for users, increased, better indexing
  • improve findability through search enginges

Semantic Markup

  • HTML5 semantic tags and schema.org
  • vocabularies that help classify page types
  • Got to 3-star just with RDF model

Results:

  • SEO, staff member went from 4th to 1st entry
  • new UI, gained capability that would not otherwise have e.g. browse by liaison area, expertise, more information rich individual profile
  • 5-star LOD
  • visualization, can download RDF, and look at graph to see who shares subjects and expertise

Next steps

  • need to check that users is helpful, especially visualization
  • testing on UI
  • auto-complete for search
  • looking at how schema.org represents a person

Lightning Talks

Linked Open Data: A New Trend of Highlighting Libraries Trustworthiness on the Web

Kaouther Azouz, PhD Candidate, University Charles de Gaulle – Lille 3

  • few academics in the field, but there is an interest in seeing linked data for cultural heritage and libraries data
  • libraries have always been a warehouse for data, and considered trustworthy organization
  • brings more visibility, etc.
  • bring awareness of problematic resources, and authenticity of the source

Unlocking the Door

Bobbi Fox, Senior Digital Library Software Engineer, Harvard University & Gloria Korsman, Research Librarian, Harvard University

  • before redesign website, gathered feedback from students
  • 5 focus groups and focused on home page, but heard about other topics
  • one topic surprised about was course reserves
    ** awkward interface
    ** LMS might not be populated
    ** prefer to search for titles than browse
    ** used OPAC more often than LMS to locate reserves, which missed many e-resources, especially if instructor posted it
  • asked students to create ideal homepage: some personalization including courses with related links, such as reserves
  • started a Course Reserves Unleashed to streamline student access to course reserves, frees up data, present data in multiple places, no authentication required to search
  • original course reserve requests was for staff, not intended for students. 10 years, but works very well
  • design considerations:
    ** forget about SOAP
    ** not to put a REST-ful wrapp with Oracle database, wanted other schools to feed into
    ** good “searchability”
    ** able to load data outside reserves tool
    ** lightweight
    ** open source (required by grants)
  • Using:
    ** Solr 4 is used as a NoSQL db
    ** course info and reserve item info in single schema file
    ** ECRU on github

Piping Hot: Little Bins in Big Workflows

Alex Garnett, Data Curation and Digital Preservation Librarian, Simon Fraser University

  • most common commands are actually small programs e.g.
  • Library example #1 PDFA: ProQuest wants PDFA submissions from now on = 5 yrs backlog. Done through gs (ghostscript)
  • Archives problem #2: GUI tool to be able to creat restrictive FTP accounts for donors using echo, read, and get
  • Library problem #3: PDF redaction. Had gray box but could still select text, so not actually redacted. Solved by using ghostscript (hard part by dumping from gs). Snip offending pages with pdftk, convert them to images with imagemagick, OCR back into PDF (minus obscured text) with tesseract and fix up the dimensions with gs again, paste back in with pdftk = 5 lines all free tools (using documentation & piping)
  • huge amount of useful space you can occupy as a barely-programmer if you’re comfortable using a terminal for problem solving
  • Piping Hot PDF slides

The User Experience Study: Student Views on the Principles of Legal Research Website of the University of Ottawa’s Brian Dickson Law Library

Channarong Intahchomphoo, Computer Reference Technician, Brian Dickson Law Library, University of Ottawa & Margo Jeske, Director, Brian Dickson Law Library, University of Ottawa

  • Principles of Legal Research Website used as bilingual online tool used by all first years
  • goals: improve experience of students by helping to build better skills, alleviate pressures faced by librarian due to heavy teaching load.
  • originally in Blackboard (WebCT), developed in dreamweaver, designed with self-assessment, quiz tools
  • website that can be embedded into LMS
  • survey results: modules were well received, appreciated having no class, convenient, ease of access (especially for return visits), but modules too long.
  • user experience: stimulating interface design, majority supported mobile version, prefer to ask questions in person, can use without strong technical skills or prior legal research knowledge
  • need to redo survey when more students can answer, need to retest for accessibility, hoping to know if other people have used any of the modules

End of the Day

After some hackfest reports, it’s the end of the day. Time for a nap before tonight’s social.

3 baby bunnies sleeping in a pair of hands

Published by

Cynthia

A librarian learning the ways of technology, accessibility, metadata, and people

Leave a Comment

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s