Mendeley – A Last.fm for Research?
Victor Henning, Bauhaus-University of Weimar,
Germany / Mendeley Ltd., London, UK
Jan Reichelt, University of Cologne, Germany /
Mendeley Ltd., London, UK
Introduction
New developments in information technology
influence the way researchers work, especially in
boundary-spanning teams and interdisciplinary
research fields. Currently, discussions of how to turn
Web 2.0 applications into productive social research
tools can be found everywhere (e.g. European
Science Open Forum 2008, Science Blogging
Conference 2008, Science in the 21st Century
Conference, “Science 2.0” related discussions in
Friendfeed, etc.), and the main questions are always –
Which tools are being used? How do these tools
improve collaboration between researchers? Does the
academic world embrace social networks for
research? Both multi-purpose social software, such as
wikis, blogs, and social networks, and more specific
services such as Twitter, Friendfeed, or CiteULike are
currently being used and evaluated by a number
researchers and academics. However, despite their
apparent helpfulness for daily research, it is even
more important to analyze the underlying principles
and concepts of these services in order to evaluate
their long-term impact and therefore usefulness for
the academic community.
Therefore, this paper aims to explore how the
principles of a well-known Web 2.0 service, namely
the world’s largest social music service “Last.fm”
(www.last.fm), can be applied to research, which
potential it could have in the world of research (e.g.
an open and interdisciplinary database, usage-based
reputation metrics, and collaborative filtering) and
which challenges such a model would face in
academia (e.g. scale, initial value proposition, and
privacy). A real-world application of these principles,
“Mendeley” (www.mendeley.com), will then be
demoed at the IEEE e-Science conference.
How Last.fm works
Last.fm, which bills itself as a “social music service”,
has managed to create the largest ontological
classification (and the largest open database) of music
in the world, by aggregating the musical tastes of its
20 million users and then data-mining it for similar
musical genres, artists, and songs. The users form a
social network that is not based on pre-existing real-
world relationships; instead, Last.fm’s network
emerges around data that describes its users’ listening
behavior and musical preferences. Can academia
perhaps learn something from the world’s largest
social music network and can its principles be applied
to research?
Last.fm works like this: Its “Audioscrobbler” desktop
software, after having been installed on a user’s PC,
starts tracking the user’s music listening behavior.
The listening data is sent to the Last.fm website,
where a profile of the user’s musical tastes is created.
Listening statistics for each song, album, artist, and
genre are aggregated and made available online. In
this way, Last.fm has created the world’s largest open
music database, comprising over 80 million songs,
accessible by everyone. The user-generated data also
lays the foundation for personalization, collaborative
filtering, and ontological classifications:
- Users can view timelines and statistics about
their own listening behavior,
- view the most popular tracks for each of their
favorite artists, and most popular artists for their
favorite genre,
- receive music recommendations based on the
song library already existing on their PC, and
- discover similar tracks/artists for every
track/artist in the Last.fm database.
The model of Last.fm applied to research
Last.fm’s service is based on aggregating the users’
existing music libraries, relationships between artists
writing songs in different genres, and the users’ music
listening behavior. Similarly, a service for academic
researchers could be based on aggregating scholars’
existing research paper libraries, relationships
between researchers writing papers in different
disciplines, and the scholars’ paper reading behavior.
Along these lines, a “Last.fm for research” would be
able to display statistics to each individual user about
his personal library, to aggregate readership statistics
about papers, authors, journals, and academic
disciplines, and to recommend interesting articles and
researchers to the user. We envision that such a tool
consists of two parts: First, a desktop application
which helps researchers manage their academic
papers (in PDF format) and anonymously tracks their
reading habits and literature usage. Second, a website
where the users can discover aggregated statistics, top
papers, trends and charts for each discipline, paper
recommendations, and introductions to people with
similar research interests. Adoption of such a service
would have a number of advantages for academia at
large, of which we will discuss three important ones.
The creation of an open and interdisciplinary
database: Similar to Last.fm’s efforts in the space of
music, a tool which aggregates metadata, tags and
article usage of a large number of researchers could
lead to an open, interdisciplinary and ontological