...
- Stony Brook – Erich has been working on the WebID authentication for another app and most of the code should be portable over to VIVO, although was done with Jetty instead of Tomcat. Can override the trust manager in JDK1.7 and implement authentication.
- Alex – might be a good topic for a focused dev call
- Florida –
- Weill - 1) Made some progress to the performance work. We noticed a lot of temp db tables being created during queries. The tool (New Relic) that we use doesn't show the exact queries that are run during the page loads. Anyone know of tools that can display the queries live during page loads? Presumably SPARQL queries get converted to SQL queries during page loads? Paul – Eliza is working on optimizing queries. Brian Lowe – you can log all the queries at the MySQL level, which can be pretty massive, but is the straightforward 2) Thanks so much for the feedback on db query cache size from last week's meeting.
- Paul – pushing to meet a deadline to provide publication data to a separate home-grown faculty evaluation tool, where they will disavow or allow any publications from 2012; using a separate tool to collect feedback to avoid corrupting data from authoritative sources, but will feed change information back to VIVO. Not yet opening up editing, and concerned about multiple entries of the same publications by co-authors – easier to figure out everything on their behalf and then ask the faculty to correct our errors, and have some people with 700+ publications. Also means that even if faculty are not interested in editing, we can provide good profiles.
- Have revisited Scopus Ids for all faculty.
- UCSF – Eric is still doing the conversion of profiles and expect to take another few weeks, and then wants to work with a new RDF to JSON standard that has come out in the past few weeks. Had been using the Epimorphics one but have found another – http://json-ld.org and https://github.com/tristan/jsonld-java
- Converting existing OpenSocial gadgets to RDF and are finding them actually simpler.
- Duke --
- SUNY Buffalo – Mark – acquired a new server and hoping to apply it to the production VIVO soon
- Scripps – Michaleen – Michaeleen – have done a huge ingest of 26,000 publications and grant data from NIH RePorter, with help from Stella; will have more news, and have been working on from an in-house EndNote library via contract with Stella; have been troubleshooting and correcting some issues in their VIVO installation and configuration
- Question: how How do you do handle author disambiguation?
- A: with Scripps has only 280 faculty can do the matching, and do comparisons against 298 faculty profiles so Stella added the names to the program which then runs a match against a faculty name field in the Endnote records; as find the program finds name variants that have not previously been recognized, they are added to the list of matches to check during ingest; problems commonly arise from suffixes (e.g., Jr., III) or special characters (e.g., publishers vary in how they handle umlauts, Muller vs. Mueller vs. Mu(umlaut)ller)
- Cornell – Tim – adding mini-reports as embedded sparql queries within individual department pages to show graduate field affiliations and research areas
- Huda – writing up usability page results; and is using the within-page sparql pages; starting on implementing additional search facets tomorrow
- Colorado – Stephen – creating new RDF for the Colorado Springs campus every other day; put out a new version for the Boulder campus VIVO with the new version of Solr that fixes a Solr bug with nicknames (see below).
- We have talked about creating a VIVO repository of testbed data and would like to pursue that soon
- Penn
- NYU
- Johns Hopkins
- Indiana
- Brown
- other
...
Access code:645 873 290