I spent time in California interviewing graduate students about their work processes. Something that stood out to me was how science and engineering students typically looked for people (rather than subject headings) during the information gathering stage. The objective was to find researchers working in particular areas and then mine their websites for additional papers. That’s exactly the approach that Scholrly hopes to improve upon.
I first came across Scholrly about a year ago when a friend of a friend liked them on Facebook. I explored and this is what I found:
“Scholrly aims to give its users, from the garage inventor to the tenured professor, a single stop for finding research connections and insights faster than ever before.”
I spoke with co-founder Corbin Pon last August and followed their development. Over the past year they’ve worked with faculty at Georgia Tech’s College of Computing to build out the idea. And in early June they’ll open up their software to beta testers.
People
What’s most interesting to me is Scholrly’s people-centric emphasis. When you search with a keyword you not only get relevant citations but relevant people as well. The goal is to let users search for people and to figure out who is important within the subject context. And not only that, but what else have those people worked on, who have they worked with, and other related connections.
The idea sparked during a conversation with a physician at Emory who expressed frustration at not being able to find people with the skill set he needed. He assumed that researchers at nearby Georgia Tech probably had the expertise, but he wasn’t sure how to identify or contact them.
This led the co-founders to thinking about how researchers are connected bibliographically and how they could also be connected through an online tool. They mapped different knowledge networks and built a search engine around the data with the objective of making the people part much more accessible.
When searching Scholrly the results are returned on two separate panels: publications & authors. While the author component isn’t too radically different, in fact many databases provide author lists/limiters in the results, Scholrly places a great emphasis on this feature. Author profiles will include career information, affiliations, publication listings, common co-authors, top publishing venues, and impact metrics. Authors are also able to claim their profile, similarly to how twitter verifies celebrity identities, and then edit/upload additional information.
Ontological Neighborhoods
Corbin often talks about scholarly neighborhoods:
“When we talk about neighborhoods, we know that there are communities of related research that are not always easy to see and explore. These researchers are reading and contributing to work from all different fields. In one project, someone might be contributing to their own self-defined field, and in another project, there could of experts for all sorts of fields. Scholrly connects papers and researchers based on the citation graph and co-authorship right at this moment, and we are developing other techniques of finding similarities. This makes it easier to identify titles and faces that appear in clusters regardless of their self described field and shows series of related papers that might suggest long term projects worth investigating. This idea of neighborhoods, we feel, better describes the structure of research than strict collections that classify work from the top-down.”
And the team doesn’t view academics are their only audience. One of their goals is to make research materials more accessible to people who don’t have access to big libraries or Fortune 500 budgets. “We really want to change the way that people find research materials” But listening to Corbin talk you realize it’s not just search behavior that he hopes will change, but the entire way people think about and approach their research process.
Oh and Google Scholar? “Our goal is to compete with Google Scholar, or replace it.”
I’ll share more about Scholrly as they progress through beta and include startup lessons learned such as their variation on the build, measure, learn methodology, their BBQ-enriched hackfests, and how they are developing and pivoting their business model. It’s an interesting product-- definitely test out the beta later this summer.
Some screenshots:
More info:
