UCSF Profiles Gives Faculty Power of Building Networks on Campus

September 17, 2010

By Maninder Kahlon and Kate Rauch

UCSF has launched a new online tool that aims to address an old problem at the multi-sited urban campus.

Say you’re a senior researcher applying for a grant. It comes back with comments. You need to improve in an area that complements – but is not at the heart of – your area of expertise. The grant reviewer is based in Washington, DC, but knows that the best person to talk to happens to work just down the hall at UCSF.

You meet up, develop a new professional friendship, and resubmit the grant successfully. It’s a true story and one that shows both how rich the intellectual environment at UCSF can be, and how hard it can sometimes be to make the connections you need.

With that in mind, UCSF has launched “UCSF Profiles,” which is part online people finder and part social networking platform, all geared for academic use. By data-mining existing public sources, the “UCSF Profiles” database has been pre-populated and is available immediately, without any researcher at UCSF guessing where to turn.

What UCSF Profiles Can Do


Most people are used to finding other researchers and potential collaborators by doing a Google search or by directly searching sites like PubMed, with fingers crossed that the scientist that they searched for is solely responsible for the hundreds of publications that are unearthed through the search.

“UCSF Profiles” improves on basic web searches, and goes even further. The tool reveals publications, co-authors, and areas of study for investigators who are experts in the subject matter of interest, for example epilepsy or a disease outbreak. 

UCSF Profiles plots out investigators that match each other based on their entire research program, not simply on one keyword, such as investigators who study Alzheimers disease in rodent models instead of patients. 

Faculty can also find out whether someone in their research network connects them to a potential collaborator. UCSF Profiles allows users to browse the network of co-authors at UCSF, even providing an interactive visual display, so they can see how many degrees of separation exist between them and the person who is searched for. The end result is a continually refreshed UCSF network of associations based on research interests, co-authorships, appointments or even office or lab locations.

The new tool is already receiving accolades from campus leaders and faculty.

“UCSF Profiles is a gateway to collaboration and a miraculous development at UCSF to support interdisciplinary research,” said Kathleen Dracup, RN, DNSc, FNP, FAAN, dean of the UCSF School of Nursing.

How UCSF Profiles Works


The tool starts with a base of professional background gathered from existing campus directories and departmental information. But the real power of the tool comes from extracting information from additional databases that hold research-related information, such as publications in PubMed. 

This information is extracted and fed through algorithms that disambiguate author names and compute weighted networks of keywords, co-authors, similar people and more. 

Many more sources of data will continue to be added to make “UCSF Profiles” as comprehensive as possible, recognizing that no one source of data fully represents the breadth of research conducted at UCSF. 

But all automatic extraction approaches need tweaking, and although the algorithms used are as good as they get and are being constantly improved, publication records may not be perfect or complete. In fact, although not necessary, faculty are encouraged to customize their individual profile pages, editing publications easily with built-in tools, and enhancing profiles with their photos and narratives. They can also choose what information they want displayed or hidden. 

Partnerships Fuel Growth


UCSF Profiles is one of several projects that the Clinical & Translational Science Institute (CTSI) is implementing that leverage new technologies and partnerships to accelerate research at the University. 

The tool is built off an “open source” product, the Profiles Research Networking Software, developed and offered for free by Harvard Catalyst (Harvard’s sister institute to UCSF’s CTSI).  UCSF was Harvard’s first partner in deploying and customizing the software, helping to test and promote the product. The universities remain close partners on the tools development and growth. 

In addition to external partners, CTSI has also made it easy for other groups at UCSF to benefit from “UCSF Profiles.” For example, the Center for AIDS Research (CFAR) now includes “UCSF Profiles” data to populate its new website.  Their faculty member pages now sport up-to-date publication lists that are automatically fed from “UCSF Profiles,” without any additional administrative burden on CFAR’s end. 

Down the road, CTSI plans on opening up the programming interface to UCSF Profiles, which will allow anyone within the community to extend the utility of the tool and/or to connect the tool to their own groups’ software. 

“UCSF Profiles” currently displays information about 2,500 faculty at UCSF. Next steps include adding fellows, residents and research staff and integrating matchmaking functionality such as between mentees and potential mentors.  Because of collaborations being facilitated at the national level, soon UCSF users will be able to expand searches to include experts at other institutions.

CTSI invites the campus community to use the tool and provide comments and suggestions for new features.  Indeed, in the final analysis, as with all social networks, the more people using “UCSF Profiles,” updating and customizing their pages, the more powerful the tool will be. Its biggest “sales force” are the users active in the system.

What Users are Saying


Below is just some of the feedback that CTSI is getting with the launch of “UCSF Profiles.”

Laurence Huang, MD, UCSF professor of medicine, chief, AIDS Chest Clinic, at San Francisco General Hospital:
“UCSF Profiles is a winner when it comes to finding people and potential collaborators, either for yourself or for your mentees. Also, as a group leader, I’m interested in including people who bring new questions and directions to enhance the group. The power of Profiles is that I can easily find who’s working in a particular field at UCSF. I can even learn who has collaborated with these people, and talk to these collaborators about the experience. This becomes really powerful.”

Dan Lowenstein, MD, UCSF professor and vice chair of the Department of Neurology, and member of the CTSI Board of Directors:
“UCSF has essentially no mechanism for identifying the characteristics of faculty members that would enable effective networking across our campuses. For example, I’m interested in the intersection of the genetic basis of common epilepsies and the determinants of pharmaco-responsiveness. At the moment, the only way I can network with other faculty interested in these topics is basically through my own institutional knowledge or by using search engines like Google. I don’t know whether the number of hits that I get actually represents a complete list or not, so this is a very inefficient and incomplete way of identifying a potential network of investigators.  With the full implementation of ‘UCSF Profiles’, I think we finally have a convenient and effective system for accessing the full spectrum of knowledge and expertise that exists among our faculty. We have a lot data, both public and internal, that we can mine.”

Ida Sim, associate professor of General Internal Medicine, and director, Center for Clinical and Translational Informatics at UCSF:
“The tool is surprisingly easy to use and especially useful for first-cut identification of people in fields I’m not familiar with. It’s complementary to meeting people in person – I can search information about someone before a meeting, or learn more about someone after a first or brief meeting. The co-author network, for example, tells me something about their level of work and scholarship, which would be more difficult to find out using PubMed alone.”

Margaret Tempero, MD, deputy director and director of Research Programs;, UCSF Helen Diller Family Comprehensive Cancer Center:
“From a program-building standpoint, I think UCSF Profiles is awesome. When I build a research program, I need to make sure I’m all-inclusive. I can see how UCSF Profiles helps identify people inside and outside of a department.  It helps me figure out which investigators I should invite to the table.”

Related Links:


Introduction to Profiles [PPT]

Profiles FAQ

Profiles Open Forum (for feedback)