Research networking and profile platforms: design, technology and adoption of
Tanu Malik, UChicago CI – treating science as an object. Need to record inputs and outputs, which is difficult, but some things are relatively easy to document: publications, patents, people, institutions, grants. Some of this has been taking place, documenting metadata associated with science. How can we integrate this data and establish relationships in order to get meaningful knowledge out of it? There have been a few success stories: VIVO, Harvard Profiles. This panel will discuss the data integration challenges and the deployment challenges. Computational methods exist but still need to be implemented in easy to use ways.
Simon Porter – University of Melbourne
Implemented VIVO as Find an Expert – oriented towards students and industry. Now gets around 19k unique visitors per week.
Serendipity as research activity – the maximum number of research opportunities are possible when we can maximize the number of people discovering or engaging with our research. Enabled by policy, enabled by search, enabled by standards, enabled by syndication.
At Australian universities have had to collect the information on research activity all along. Some of it is private, but some is public and the University can assert publication of it. Most universities have something, but lots of different systems.
Only a small number of people will use the search in your system. Most will come from Google.
Syndicating information within the university – VIVO – gateway to information – departments take information from VIVO to publish their own web pages. Different brands for different departments.
Syndication beyond the University – Want to plug into international research profiling efforts.
New possibilities: Building capability maps. How to support research initiatives. Start from people being appointed to start the effort. Use Find An Expert to identify potential academics. Can put together multiple searches to outline capability sets. Graphing interactions of search results.
Leslie Yuan – Clinical and Translational Science Institute – UCSF
The Profiles team all came from industry – highly oriented towards execution. When she started they wanted lots of people to use, so how to get adoption? If you build it, they probably won’t come. Use your data and analyses to drive success with a very lean budget. In four years went to over 90k visits per month. Gets 20% of the traffic of the main UCSF web page.
1. Use Google (both inside and outside the institution). Used SEO on site. 88% of researcher profiles have been viewed 10+ times. Goal was to get every one of researchers to come up in top 3 results when they type the name in. Partnered with University Relations – any article that the press office writes about a researcher links to their profile.
2. Share the data. APIs provide data to 27 UCSF sites and apps. Has made life easier for IT people across the university, leading to evangelization in the departments. Personalized stats are sent to profile owners – how many times your profile was viewed within the institution, from other universities, from major pharmas. People wanted specifics. Nobody unsubscribed. Vanity trumps all. Research analytics shared with leadership. Helped epidemiology and biostatistics show that they are the most collaborative unit on campus.
3. Keep looking at the data – monthly traffic reporting, engagement stats (by school, by department, who’s edited profile, who’s got pictures), Network visualizations of co-authorships.
4. Researcher engagement – automated onboarding emails – automatically creating profiles, then letting people know about them as they come on board. Added websites, videos, tweets and more inline. Batch loaded all UCTV videos onto people’s profiles, then got UCTV to send email to researchers letting them now. Changed URLS – profiles.ucsf.edu/leslie.yuan
5. Partnerships – University Relations, Development & Alumni, Library, UC TV, Directory, School of Medicine, Center for AIDS research, Dept. of Radiology. Was able to give data back to Univ Relations on articles by department or specialty, which they weren’t tracking. Automatic email that goes out if people get an article added.
Took 8 or 9 months of concentrated conversations with chairs, deans, etc to convince them that this was a good thing. Only 7 people asked to be taken off the system. Uptake was slow, but now people are seeing the benefit of having their work out there. 6 people on her team have touched the system in some way, but it’s nobody’s full-time job.
Griffin Weber, Harvard – Research Networking at the School, University, and Global Scale
Added passive and active networking to the profiles system. Passive network provided information that people hadn’t seen before, driving adoption, active networks allowed the site to grow over time. Passive network creates networks based on related concepts. Different ways of visualizing the concept maps – list, timeline, co-authors, geography (map), ego-centric radial graph (social network reach), list of similar people
Different kinds of data for Harvard Faculty Finder – comets and stars discovered, cases presented to the Supreme Court, classes taught, etc. Pulled in 500k publications from Web of Science. Derived ontologies in 250 disciplines across those publications using statistical methods.
Direct2experts.org – federated search across 70 biomed institutions.
Faculty affairs uses Profiles to form promotions committees, students using it to find mentors.
Bart Trawick, NCBI – NLM – Easy come, easy go; SciENcv & my bibliography
NIH give $15.5 in grants per year. Until 2007 didn’t have a way of seeing what they were getting from the investment. Public access to publications mandated by Congress in 2007. Started using MyBibliography to track. Over 61k grant applications coming in every year, just flat PDFs.
About 125k US trained scientists in the workforce now. Many have been funded by training grants. Want to see how the scientists continue their career. Over 2500 unemployed PhDs in biomedical science.
My NCBI Overview – tools and preferences integrated with NCBI databases. Connected to PubMed, genomics, etc. Uses federated login (can link google accounts e.g.) Can link ERA commons account – pull in information about profiles, grants linked.
My Bibliography – make it a tool to capture information and link grant data to publications. Set up to monitor many of the databases that information flows through. End result of public access policy is that all NIH-funded research publications get deposited in PubMed Central. MyBibliograhpy lets scientists know if they’re compliant with policy. Send structured data back out to PubMed, allowing searching by grant numbers, etc.
SciENcv – released second version this week. Help scientists fill out profile – each agency has their own biosketch format. SciENcv is attempt to standardize that. NIH set up, working on others, NSF next on list. Wanted to make it easy for researchers who are already funded and using MyBibliography. Data exists out there – would like to get to a point of reuse of data for grant reporting. Added inputs – ORCID, eRA Commons (used to manage grants), MyBibliography. Grants.gov requires biosketches in PDF. Can export from SciENcv in pdf to grants.gov, with rich metadata attached.