11:00 am Eastern time
Agenda
- Updates, issues, questions or feedback from sites
- Representing that a Journal is open access (http://issues.library.cornell.edu/browse/VIVOONT-433)
- Developing a D2R-based data ingest option for VIVO for those with relational DB skills and resources.
- We'll review the goals of this effort, progress so far, and Indiana University and Cornell's current work on the schema for the relational DB middle layer.
Call-in Information
1. Please join my meeting: https://www1.gotomeeting.com/join/956365752
2. Use your microphone and speakers (VoIP) - a headset is recommended. Or, call in using your telephone.
Dial +1 (646) 558-2103
Access Code: 956-365-752
Audio PIN: Shown after joining the meeting
Meeting ID: 956-365-752
Notes
Site updates
Scripps: Migrated to 1.4 and that went fine, currently working on some data cleanup.
Weill: Effort to use Scopus as a source for authoritative publication metadata - have done a first pass matching scopusId to author, still needs some manual matching to be done afterwards… - maybe they will have a staging DB to let authors verify accuracy before publishing to VIVO. Working on documents about governance and policy.
University of Florida: Migrated to 1.3, working on data cleanup. Upon moving to 1.3 deleted about 30 UF extension classes for Organizations and replaced them with the Internal Institutional Class for organizations (UFOrganization). This worked well; we discussed moving towards the direction of a high level Internal Institutional Class that could be asserted for, e.g., Person also. Nicholas intending to make any 1.4 updates to the ontology diagrams he created for 1.3, and to not include (or have versions that don't include) UF extension classes and properties.
University North Texas: Building up team of 4 or 5: students, MDs, Kathryn, Cliff. Cliff prototyping and looking at data sources such as Google Scholar and also an internal faculty profile system. Doing some work with zotero. Bill Moen is attending iConference. UNT quite interested in the ingest-from-relational-DB effort.
IU: at version 1.3, doing data cleanup.
Representing that a Journal is Open Access
Paul (Weill): want to show that VIVO can be used to generate interesting reports in this area. UNT also very interested in Open Access designation; have just created an institution-wide open access policy that Bill Moen championed. Scripps: this sounds useful for the future, but not high on the priority list at the current tine. Notes on modeling requirements recorded in Jira issue VIVOONT-433
Discussion about middle-layer DB schema
The goal is to provide a tool to facilitate ingest of data into VIVO for those with relational DB skills and resources. We would provide a well-documented relational DB schema and tools to generate VIVO RDF from the information in the tables. A site would then have the option of populating this DB, and from there could use the provided tools and framework to automate the transfer of the data into VIVO. IU plans to investigate both D2R and XSLT implementations for generating VIVO RDF from the middle layer DB.
Person and EducationalBackground schemas
Robert, DBA at Indiana University: the schema for this DB should be intuitive to a typical DBA, and not sure the Cornell-proposed schema is intuitive. The title of the table should convey what kind of information the table contains and believes that, e.g., "ThingConceptLink" is not a good table name.
| Panel |
|---|
Ying: Would rather put more burden on the mapping to do the transformation of form rather than have the tables closely reflect the ontology classes, feels that will make it easier for people. Mentioned that the DB tables should be normalized, at least to some level, but did not believe that either of the schemas is not normalized.
| Panel |
|---|
Jon (via e-mail): Designed and proposed these modifications to the schema because would like to have the relational model stay as close as possible to the VIVO ontology, in form and in the naming of fields to match properties – we want people to assimilate information about the VIVO ontology, and designing an optimal structure in the RDB for performance is not the goal.
| Panel |
|---|
Stella: believes that having a schema closely following the ontology in naming and structure and that follows simple conventions in representing classes, object properties, datatype properties, and the one to one, many to one, and many to many relationships in the ontology, will result in an infrastructure that is more easily designed, more easily reviewed for correctness, more maintainable, and easier to debug. Since it will be a relational DB schema, we would expect that a person familiar with that technology should be able to understand and populate the tables, even if they do not think it is the most intuitive design.
Kathryn: interested in reviewing both schemas and giving feedback.
Discussion to be continued.