Mendeley is a service that provides a tool for researchers to easily comprehend the nature of the readership surrounding a particular academic paper or collection of papers. Not only does it determine the credibility of material by means of social proof it also provides reader recommendations which makes it easier to collect and collate a library of material on a given subject.
It was founded by three Phd students in Germany. One of them, Dr. Victor Henning, wanted to be able to look at a paper and see how the citation network connected that paper to the other papers he was researching. It was from this initial curiosity that Mendeley evolved.
We spoke to Ian Mulvany the head of New Product Devlopment and asked him, how that original inquisitiveness led to the discovery of the problem which they set about solving.
“From the end user point of view, for the scientist, one of the problems they have in managing their information is that it is a bit of a mess. The academic articles that they access, that they publish, that they download from the web mostly come in the form of PDFs. Managing the metadata around that is painful.
“It’s like you take a highly structured piece of content and you put it through a sausage factory and you end up with this digital object which has practically no metadata associated with it as far as the end user is concerned.
“But there’s another problem which is even bigger. It is not just about knowing the smaller piece of information that you are interested in. The bigger problem is, how does that fit into the bigger global picture of the entire academic content and the entire world of academic literature? How does the article you are reading connect to an article that someone in a different field is reading? How do you know about the other people who are reading the same kind of things you are reading about?
“So what we have done at Mendeley is create a single tool which sits on your desktop and helps you manage your individual articles. We mirror all that activity into the cloud so we can see what people are reading right now and how many people are reading articles on a particular topic. We can also see how many people are reading a particular article. Our vision is to use this user activity around their local usage of PDFs to create connections between researchers.
“The more people that use it, the more crowd sourced the value we generate out of people’s usage. Using our client application as a base we have created the world’s largest search catalogue for academic papers which also gives you social information around the usage of those academic papers.”
How does this compare to what Google is doing?
“Google have a product called Google Scholar which indexes content. But what Google don’t have is access to the individual collections of articles that someone has decided that they are interested in.
“They provide many of the search services and discovery services but they don’t know what people are actually reading. They might know what people have browsed to but once someone has actually got their article Google doesn’t know anything else about them.”
What is the role of the academic publishers in this area?
“They have a lot of really rich information that they could use to generate interesting services on top of the content for scientists. Publishers have full access to the content. If they are looking at their server logs they know who is reading the content. They know who the authors are. But there has been little interest from the publishers in doing that kind of thing.
“From the point of view of the academic publishers it is purely a volume business. They just need to get as much volume out the door and sell subscription packages to libraries. If you look at the bigger publishers it is no longer enough for them to be a content paywall. They need to be providing services rather than just content.”
All networks, social or otherwise, have privacy issues. How do you handle that at Mendeley?
“We anonymize the information around who is reading what paper. The benefit to a user for using our service is if they synchronize their data with our client based service they can keep all of their information synchronized between multiple machines. As soon as the information passes through our servers we know what the reader information is but we anonymize that in the online catalogue.
“If you want to be even more private around your data you can stop information from going up to our cloud but then it won’t sync between the different machines. You can contribute to this online catalogue of usage information without exposing your own personal collection. We don’t identify paper X is read by reader Y. We say, paper X is read by X number of readers. This percentage are in America this percentage are in Europe. This percentage are professors, this percentage are undergraduates.”
Could you give us a brief example of how it might work for a user?
“When people create an account with we ask them for their academic status; undergraduate, graduate, Phd student, professor and so on. We ask them what areas of research are they interested in. We make the aggregates of those attributes around the papers available through our API.
“You could ask the following question through our API, “In the last year, what has been the most read by undergraduates in biological sciences in North America?” Presumably, that is going to be quite a different paper than the most read paper by people who are senior professors. These are the kinds of questions we are enabling people to answer that nobody else is really doing at the moment.”