Yesterday I attended a presentation at the Weill Cornell Medical College on VIVO, a system that can aggregate information about research and scholarship from a variety of data sources. Although it comes out of the biomedical research community, VIVO can theoretically supply this same functionality to any academic discipline. In addition, because information in VIVO is typically derived from an academic institution’s records, it is generally more trustworthy and accurate than information in a system like LinkedIn or Academia.edu, where an individual can edit their own profile and so could misrepresent themselves and their expertise.
VIVO is tied into a number of other systems and technologies. First, it’s built on semantic web technologies, specifically
Linked Open Data, which means it’s more extensible and flexible than other systems. It also relies very heavily on other standards, including Open Researcher and Contributor ID (ORCID), which provides a unique identifier for researchers; structured data in PubMed; and the Digital Object Identifier (DOI) system.
Although VIVO has primarily been designed to present information about researchers, rather than detailed description of research resources, the project has been working with the Consortia Advancing Standards in Research Administration Information (CASRAI) to develop a technology-neutral transfer standard for research data. They hope to use this standard as a bridge between VIVO and eagle-i, a project based at Harvard that focuses on gathering and presenting data about resources used in scientific research, including publications, protocols, software, databases, specimens and instruments.
At the end of the informational session there were a few demonstrations from institutions that have implemented VIVO. Two people from SUNY Stonybrook showed off SUNY REACH, which aggregates and presents information from a number of SUNY schools.
Yin Aphinyanaphongs from NYU showed off their “Find a Collaborator” application, which is a search layer built on top of VIVO. Although this application can only be viewed within the NYU network, the discussion of their search design was fascinating. Instead of starting with an algorithm and then tweaking it based on hypotheses, they started by asking experts in the field what kind of results they would expect to see for a search of other experts in their field and then designing algorithms that would return results that matched this “gold standard.” It was also interesting to hear his take on use of the system, which was “I want people to spend as little time as possible on this website,” the idea being that the less time they have to spend looking for someone to work with, the more time they have to research and collaborate. While this makes good sense, there are also implications for how one analyzes use statistics for a discovery system; short visits to a website are usually considered negative, but perhaps they are a positive indicator for a discovery system.
Other institutions, including Weill Cornell and Brown University, also demoed or discussed their implementations of VIVO.
Overall, this was a very interesting meeting, and while we probably don’t have any immediate application for VIVO here at RAC, it’s good to have a greater understanding of the technologies that other institutions are using for search and discovery, and the standards landscape in general.