Knowledgespeak: So weíd like to welcome Victor Henning, co-founder and CEO of Mendeley. Thanks for coming for the Knowledgespeak interview.
Victor Henning: Thanks.
Knowledgespeak: Weíll just dive right in here. We know youíve only been at the Frankfurt Book Fair for a few hours. Can you briefly tell us about yourself and about your organization, Mendeley. When and how did the inspiration to create Mendeley come about?
Victor Henning: The inspiration for Mendeley came about because I was a grad student myself, roughly 5 years ago. I started working on my Ph.D., and I realized that my cofounder who was a Ph.D student in a different field, a different university, had the same problems as I have, which was an efficient way of managing information. So, we started wondering about writing software that would automatically extract information from PDFs, like author, title, volume, journal issue - basically turn your collection of PDFs into a structured database for you. And the second conceptual leap was if we had software that can do that, and if we network all the people using the software, then we can aggregate this information and thereby make it much more meaningful because to each document in the database, you add the social information about who it is that is working with the document; in which discipline is the document popular; how many readers does it have; and are the readers undergrad versus Ph.D. students versus faculty; geographic distribution; interest over time So all of these social matrix that you can add. That was the basic idea for it.
Knowledgespeak: Great. Thanks for that. How receptive has the publishing industry, especially STM publishers, been to the Mendeley concept?
Victor Henning: Very good, so far. So thatís been good news for us. When we started, and I think still, our focus is on providing a productivity and collaboration solution, not so much a publishing platform or a distribution platform yet. However, one of the things that I think publishers have always seen in us, and they basically have been in talks with us since we started the company, was because we have all of this information about which people are interested. Based on their existing document collection, Mendeley can become a very good platform for reaching people in a very targeted fashion.
So, I always thought about it in terms of not replacing the subscription business that the publishing industry has traditionally pursued, rather offering avenues for additional monetization of content by selling it to individuals who might be interested but where we know that they donít have the document yet and donít have access through their institution, for example. Apart from that theyíve also been very supportive to the conference that we actually host in London every year. The Science Online Conference where we discuss various science and scientific collaboration and science communication, and also, to a certain extent, scientific publishing going in the future. So there is some interest.
Knowledgespeak: Fantastic. Thank you. In what ways is Mendeley different from other desktop reference managers, such as EndNote. And what are the key differentiating factors that set Mendeley apart from its competitors?
Victor Henning: I used to be an EndNote user myself. Its great for what it does. Itís a reference management tool with a plugin for Word where you can generate citations. But the thing that I always found cumbersome was keeping the database up-to-date and current. I always used to forget to enter references once I had the document and then I wasnít connected to the PDF and couldnít do a full text search. I was using yet another tool to read the PDFs and annotate them.
So, Mendeley basically takes the hassle out of that. You can just basically point it to a couple of folders on the hard drive and whenever you store a new PDF there, itíll automatically get added to Mendeley and then itíll extract all of the information and turn them to a database for you. And we do have plug-ins that allow you to cite the documents. So, essentially the same sentimentl is there. In addition to that, we have the annotation and full-text search capability.You can view your PDFs in full-screen, add little sticky notes and highlights, and even better is the collaboration functionality. Because you have a Mendeley rapid count, you can synchronize your library to the web; you can invite other people to share folders of documents, which is great for lab collaboration; and I would say its very hep in the sense that you get the news feed of whatís happening in a group. So if you know the Facebook news feed then you can imagine what it looks like. That news feed is available both on the web--so if youíre pointing the browser to it, you can access it from anywhere--but also it is pooled into the desktop interface itself. So you can actually benefit from having the best of both worlds - you have the user experience, the fluidity, the quickness of having desktop software combined with the interaction and collaboration and the news feeds and updates of what is happening in a group through web collaboration. Plus we do have mobile apps as well. So your library is basically synchronized across multiple machines: your home machine, your office machine, your iPhone and iPad and online. So its very very useful for that.
Knowledgespeak: Great. Thank you very much. This question really follows on to what you just shared. No doubt social networking is becoming a mainstream component of scientific research. We know that social networking is an aspect of Mendeley. Where does Mendeley integrate with other social networking sites, exclusively for scientists for example, Laboratree or ResearchGATE.
Victor Henning: I think, as you correctly pointed out, social networking is an aspect of Mendeley and I think itís an important distinction that we donít treat social networking as the end in itself. To us itís more like needs to an end, and we believe that what scientists really care about is research papers and research data as the social objects. So the analogy would be with Flickr - its not a social network - people form groups around images they find interesting, and so it is on Mendeley. One of the things that really surprised us was how many documents people were uploading to Mendeley. The current count after about 21 months of being public--forty two million documents have been uploaded. So itís about 300 thousand a day; And this is where groups form around it. We havenít yet integrated with other social networks for researchers because we think what people really want is to collaborate on the objects, like research objects. Having said that, we do have an API now for roughly 6 weeks. If other platforms want to integrate with Mendeley, they can now actually do that. And we see the first tools, both publishing platforms and new applications, being developed on top of the API. So one example would be ReaderMeter, a site that measures scientific impact of authors based on their readership on Mendeley.
Knowledgespeak: Thatís great. Thanks a lot. Can you briefly talk about the recently launched ReaderMeter, which measures research impact based on readership-based metrics instead of citations? Do you feel it is a more meaningful way to measure the importance of a paper and/or a journal? If so, why?
Victor Henning: One of the things that did frustrate me when I was still working as a researcher full-time, and actually I still have a couple of papers under review--so I probably should be careful of what Iím saying--is that it simply takes a tremendous amount of time to get papers published and not just getting through the peer-review process, which you know, can take lots of iterations and several months, but then once the paper is accepted, it might be another year till the paper is in print and then maybe another year for citations to be measured--actually in Thomson Reutersí case, the impact factor looks at the citations of the past two years. So one thing that always struck me was that libraries make purchasing decisions based on the impact factor. Scientific careers get determined by whether youíre publishing in high-impact factor journals. Grant funding is the same. All of this is based on a metric which is backward looking.Ėthree years. One of the advantages of social networks or collaboration tools that crowd source information, like we do, is that you can basically peer into the workflow of researchers as it happens. So Mendeley, for example, has this enormous database of 42 million papers, anonymously aggregated. Each document comes with information about how often itís been read and by whom. Though you couldnít look up a document and say person X has read it, but you get the information that this is a paper thatís popular right now in biological sciences for a senior faculty at top institutions in the United States. I think that if you understand citations as a proxy for this paper to influence someone enough to cite it, then I think readership would be another proxy for that. On top of that, in Mendeley, because people tag documents, they apply some semantic context to it so if this paper is about ABC - itís good, itís bad, is the paper being read in detail, is it just scrolled through, how often is the paper highlighted, annotated. Actually if you look on any of the user profiles on Mendeley, like my own, there is little graph now that shows how many readers the papers that I have published have on Mendeley for each of the papers. For authors itís also a nice tool to see how their content has been consumed by their peers.
It occurred to me that I didnít answer your original questions about the ReaderMeter . ReaderMeter was developed by Dario Taraborelli, who is a postdoc at the University of Surrey, and he has done quite a bit of research on social peer-review based on social bookmarking tools - does that say something about the quality of research. He basically pulls data from our API. If you enter any author name, it will show you all of the publications of that person that are in the Mendeley database - shows you how many readers each document has and then calculates an HR reader index for that person, shows you the co-authors and the geographic distribution of the readership. So itís a very nice tool that we actually hacked together in a couple of days using the data that we aggregated. So I think thatís really exciting.
Knowledgespeak: That is exciting. And our last question for today--please share with us any interesting events or user trends you have experienced this year at the Book Fair. We know youíve been here only a little while, but so far has anything jumped out at you?
Victor Henning: Actually it did. One of the reasons I came here was because I was invited to a panel called ďDo social networks threaten academic publishing?ĒItís kind of a strange thought because weíre not trying to replace academic publishing. Nonetheless, there were some people concerned about what if,suddenly the readership on Mendeley becomes the metric by which documents are measured, and also what if the filtering function that publishers have is taken over by peer-judgments in social networks. So, instead of submitting my paper to a journal with an anonymous peer-review, where the reviewers might be competitors of mine, I could just put my paper on say Mendeley or another social network and have my entire peer-group review it and have immediate access. And what struck me, however, at the panel was that most people actually thought about the opportunities - more about how can we integrate this into the workflow, how can we make the content more highly available and more useful to people. So despite the fear-inducing panel title, I think, most people were very receptive to this concept and that might have been different 2 or 3 years ago. So that was one thing that struck me.
Knowledgespeak: Well thanks so much. We really appreciate all your comments and insights and for coming out and doing an interview today.