February 23 LD4L Workshop breakout session: New Uses & Use Cases for Linked Data

facilitator: Tom Cramer

Table 1

• Collection management: shared collections, ILL, acquisitions
• University press: identify authors, topics, readers
• Usage data :: Publications :: profiles :: faculty evaluation
• Data Repositories publishing LOD to facilitate data reuse
• Trending (what's happening on Twitter, Reddit, etc.) w/ links to LOD from LAMs
• componentized calls: "data the web wants" for thumbs, licensing, attribution, TOC, etc.
• help drive people to OPEN data
• Linked-data based cataloging, description
• give researchers a platform for linking data
• crowd/club-sourcing links, topics, terms, etc.
• citation mapping & mining

Table 2

• Find MY library resources in ANY search engine
• use existing use cases to guide implementations (don't need to make new ones...)
• show business / $ value
• Think about new users: K-12, genealogists, tourists, etc. 
• Find my library resources reused in the web via mashups (exhibits, tour guides, a la IIIF, etc.) 
  • let users lasso content
• Educate / train others in linked data / web structures (other depts on campus, e.g.) 
• "old cities online", pulling LOD from memory organizations with historic data about cities

Table 3

• Consuming & publishing use cases
• Non-latin script and language support
• Recommender services, like Facebook / Linked-in reading lists, cross-linked with local library holdings
• Link paths to best access for information
  
  • around paywalls
  • best version for a device
  • local holdings
• Preferred/ Trusted provenance filtering / views
• "follow the data trail" – track back to see who made what assertions, changes, in parts of graph
• Export / machine APIs for reuse elsewhere
• Capture curation/exhibits/storytelling for reuse
• don't collect–just refer
• Preserving the graph
• Automated pub quiz

Table 4

• Talk to users, not librarians
• do things with value
• do new kinds of catloging
• Talk to non-users
• Multi-lingual support
• New types of queries / browse / interactions made possible by LOD
  • make it shiny – have compelling UIs
• capture annotations from collaborative study for reuse (crowdsourcing...)
• talk to machines
• talk to 19 year olds (younger users with different use patterns)
• mine known library taks, such as known item search
• convert opaque RWO to ones with internal links
  • video from a single file to one with video & audio separate, e.g.
  • books with marginalia into text w/ annotations