Hyatt Regency, Indianapolis, IN June 18-11, 2015. 400 attendees. Mostly librarians and developers. Strong Duraspace presence in projects, staffing, sponsors.
Fedora will over training in October. Three days. $1000 for industry, $600 for academic, $400 for Duraspace member. Donated space at a university. No instructor compensation. Must have 10 students or cancel. Make $10K twice a year.
Monday
Duraspace reception. Hundreds of people. Spoke with conference organizer for next year – Trinity College, Dublin, Ireland. High Performance Computing background. Learning about libraries and repositories.
Tuesday
40 minutes of intros
Keynote Kaitlin Thaney via Skype
Mozilla Science Lab. Help researchers leverage the power of the web. Code, community, data literacy. Our current systems continue to create friction. What is web enabled research? Access to content, web native tools, rewards for openness, push for reuse/transparency. Research, social capital, capacity. Content, context, access. Code as a research object. In figshare. OpenBadges. OpenResearchBadges. Designed icons for Credit taxonomy. Sharing detailed research data increases citation rate. Mozillascience.github.io/studyGroupHandbook. OpenScienceTraining. Mozilla Fellow for Science. Mozillafoundation.org
W3C Linked Data Platform
90% of LDP is REST. As of February 2015, a W3C standard. RDF in turtle or JSON-LD. Containers described in RDF. Post content (image) to the container URI. Supported in Fedora 4.
Peter Sefton
Ozmeka. Omeka. Lightweight. Easy to hack. RDF aware. Linked Data != RDF. Ozmeka on Github.
Rosemary LeFaive UPEI
Upei Frbroo and CiDOC CRN. Linked data. Ontology. Cultural heritage materials. CIDOC 1999. ISO standard since 2006. Current version 6.1. Feb 2015. 89 classes. 285 properties. Object oriented, event centric. FRBRoo extends CIDOC to bibliographic representations. Works of various types. Used by the British Museum and others.
Orcid panel
Confusing panel regarding relationship of repositories and orcid. The answer is straightforward -- support orcid as an optional attribute of contributors.
Laura Paglione. Technical Director, orcid
Vertical repositories. Hierarchical. Aggregation. Distributed repositories. Projects cut across. Orcid not a profile. It's a bunch of links. Displayed as a profile. Employment and education data. Works, employment, funding, education, peer review. Most publishers are now asking for orcid during manuscript submission. FundRef and CrossRef. Orcid notify services. Researcher permission to update. Collaborative with share notify.
Wednesday
Anurag Acharya. Google Scholar
Scholar groups versions of work. Determine article metadata. Crawlers don't use Ajax, post, knit chapters together. URL plus text. In one GET. Use high wire press metatags. Citation_pdf_url. Don't reorder/change the author list. Don't mess with publication date. Author/contributor blurring. Scholar tries to auto identify sites with problems. Provide year month browse. Provide new articles. Provide metadata. Use citation_pdf_url. Author list exactly as in the article. Eprints. Very good. Dspace. Recent releases 1.7. Good. Digital Commons. Often well indexed. . PURE. Popular in Europe. No Machine readable metadata by default. OJS. Good defaults for indexing. Google Scholar focused on articles. Not interested in data or other artifacts. Schema.org is not helpful. Orcid may not be correct. Include author string as received. Need detail for redundancy
John Kearns -- Symplectic
Harvest only identifiers from Orcid. Go to the source to get the data objects. Don't use Orcid data. Just the list of doi, etc. Imperial created 3500 Orcid with extensive communication. 400 already had an Orcid. 240,000 works submitted from symplectic to Orcid. Jisc report.
Tib Berlin
Using archivematica. Packages. IRods storage middleware. Fedora data management. Islandora for access.
Databrary, NYU
Good talk from NYU regarding video in archives -- working with researchers. Curating data is labor intensive -- video not self describing. Complete solution -- angularjs front end, play/scalar middleware, postgreSQL. Databrary. GitHub.com/databrary
The Portland Common Data Model
Hydra -- rails application. Pcdm: collection, object, file. Should make connection with VIVO-ISF.
Discussions
FAO UN regarding VIVO in Central America, focused on agriculture
Gottingen, DE. Possible Duraspace event in Europe in the spring.
Graham Triggs. VIVO development and features.
Alex Viggio. Community building. Roadmap process.
Mark Hahnel, Figshare. APIs. Grey literature.
Zenodo. DOI. CERN.
Piotr Hoebl. Northwestern architecture, relation to VIVO.
Smithsonian. VIVO and CIDOC for cultural heritage. Karma for ingest.
David Wilcox. Fedora training programs
SHARE Notify. Open Science Framework and Share Notify as a source for VIVO. 41 sources including CrossRef, PubMed, FundRef. Sources will grow. They do some data improvement.