Science and Web 2.0
Ian Mulvany, Nature Publishing Group
I hope that the the take home message from this short talk will be that web technologies could have
important role for science, that they are in their very early days at the moment, and that over the
next few years the outlook is good.
I’m going to tell you a little bit about myself, about what Web 2.0 is, and why Nature is interested in
After that little overview I'm going to to run you through a short demo of some of the things we
have been developing at Nature for scientists.
My name is Ian Mulvany, and I’m a product development manager with the Web Publishing Group at
I began a PhD in Astronomy, but then moved into academic publishing. Last year I began working
with this group at Nature because I am really excited about the opportunities the web has for
I’m just telling you this because you are mostly a Cell Biology audience, so I’m interested in hearing
from you what tools you use and what sorts of things we are doing right or are doing wrong from
If you want to keep up with what our group is doing we have a public blog at http://
What is Web 2.0?
Historically the term was ﬁrst used as the name of an O'Reilly Conference held in 2004 that
discussed the post
dot com bubble reemergence of internet business.
What has this got to do with science?
We will see that the deﬁning methods that new tech companies use may be applied to issues of
concern to scientists.
At the heart of the Web 2.0 approach is getting and using data, so for a scientist it should make
upcoming.org and EVDB
domain name search engine optimization
cost per click
The meme of Web 2.0 was inﬂuenced by comparing pre dot com bubble companies and post
dot com bubble companies.
What is the difference between the list on the left and the list on the right?
Let’s take the example of Brtiannica vs Wikipedia.
The information in Britannica is centrally controlled. It has a relatively small number of contributors.
The workload per contributor is high.
Wikipedia is open to anyone to contribute. A collaboration of 1000’s can lead to a work of equal
a more centrally controlled method.
Britannica’s revenues decreased from 650M to 50M over a 10 year period!
The new sites make it easy to add information and use that information to
answer or solve problems for people.
easy plain text, emails hyperlinks
academic papers semantic web
hard mining easy
Let’s look at formats for data in the space of how easy it is to create,
and how how easy is it to mine for interesting information.
Plain text is the easiest to create, but is very hard to do data-mining on.
Text that has been rigorously annotated (semantic-web stuff) is very easy to do data-mining on,
but it is hard to get people to make this sort of data in the day to day activity of their lives.
(One of the things that helped wikipeida be success was a tool that enabled people to easily add
articles that got converted into nice looking web pages)
Unfortunately academic papers are really hard to write, and usually are only available in pdf, the
worst of both worlds.
Hyperlinks, page views, tags, and possibly academic citations, are easy to create
and are easy to do data-mining on. I’ll show you how some people have used these to build
of the web today.
Google created their search engine by looking at hyperlinks.
If lot’s of pages linked to one particular page then that page is probably important.
(the big red ball here)
If that page has a link, that link is important. (link from the red ball to the orange ball)
This is a mirror of the academic citation system. Sergi Brinn and Larry Page were PhD students at
when they founded Google.
You can see a conference paper that they wrote about their search engine here:
Amazon use page views and a database of user purchases to ﬁnd things you might like.
Again, here they are using data that they get for free from people using their site.
The last two examples used basically static sets of data. The data is being updated, but real time
information is more or less not required to ﬁnd the best web page on a topic or to determine
a buying recommendation. (searching for news is very different.)
Google Ad Sense is different.
Putting ads on a web page does need almost real time information. Where has the person looking at
this page been, where are they physically located? How much time have they spent looking at
The more accurate your matching based on behavior, the more money you are going to make.
We have seen how mining user provided data can help to solve problems about information
on the Internet.
Are there any problems in Science that could help with these kinds of approaches?
I think there are lots, and I think they break down into a few different categories. Let’s have a look
at a few here.
- information management for the individual scientist
- communicating with the public
- mining data
On this slide we have our researcher scanning his favorite journal
But the proliferation of journals has led to a problem, what to read?
How do we get the reader back in control?
Aside from journals, there are also lots of other places that the scientiﬁc conversation has
Discussion Groups and Mailing lists contain a huge amount of information from
from snippets of computer code, to long discussions about topics.
Mark Mail, from MarkLogic, have a site that mines this information. Here we see
a comparison of a search for FORTRAN vs a search for Java.
At the moment these kinds of archives are mainly relevant in the computer science area, but
these kinds of conversations are going on all the time in every ﬁeld.
Science blogs represent another medium where science discussions are happening.
These are just some of the sources that might begin to load up your reading inbox, but as I
the web also offers great opportunities to communicate and share your scientiﬁc excitement with
Bugscope shares a scanning electron microscope with school classes across the world
The Faulkes telescope does the same with an astronomical telescope.
There are some solutions to these problems that are out there.
The Chemical Blogspace, run by Egon Willighagen from Wageningen University, automatically
collects articles from blogs about chemistry and looks for the most popular articles.
Jean-Claude Bradley from Drexel university does open notebook science.
You can see the data and lab notes for each experiment as it is done is his group through
their blog http://usefulchem.blogspot.com/.
A part of his motivation is that all of the data is never gets through to a published paper
can also be very valuable for the community.
In arguments about precedence where there is an existent trial of discovery getting
scooped is going to be harder.
There do exist senantic-web approaches to science. The Crystal Eye project from the University of
Cambridge is trying to automate the recognition of crystallographic data in academic papers.
Science Web 2.0
Though not exactly the same, web 2.0, Open science and the semantic web work well together
and they share some common traits, namely sharing and openness of information.
This is leading to a brave new world in the space of scientiﬁc conversations.
So why are nature involved in these kinds of non-journal initiatives, and what are the initiatives that
we are involved with?
• quot;It is intended, ﬁrst, to place before
the general public the grand results
of scientiﬁc work and scientiﬁc
• quot;to aid scientiﬁc men ... by affording
them an opportunity of discussing
the various scientiﬁc questions that
arise from time to timequot;
This is Norman Lockyer, the ﬁrst editor of Nature, and these are snippets or our mission statement.
One and one way of looking at the whole issue is that journal publication is only one aspect of
communication in science.
As we have seen there are a lot of new channels for communication, apart from journals
and if Nature wants to remain relevant then we have to engage with these new emerging
We have a group in Nature, web publishing, whose goal is to be experimental and try to both keep
up to date with what is going on out there in terms of new developments, as well as trying to create
new tools for researchers.
Nature Web Publishing
The main products that we have developed so far are
- database gateways
- OTMI (open text mining interface)
- nature network
- nature preceedings
Nature Network is a place for hosting discussions and forums about science related topics.
Second Life Nature
- lectures in second life
As my colleague Jo who works on second life likes to say, SL is a platform that offers a lot
of potential, however not many people know what the best way to use it is. We have been hosting
couple of scientiﬁc projects there,
UCL CENTRE FOR ADVANCED SPATIAL ANALYSIS
Drexel Chemical Reactions
Artiﬁcal eco system, some of the creatures escaped our island and were found in the wild,
and I'll be happy to answer questions about these later, but
But so far the most successful thing we have done is use it as a location for hosting talks for the
Scintilla is like the Chemical Blogspace in that it aggregates content from over 700 science related
You can tag, share and rate stories that you like on this site.
Precedings is a preprint server for the life sciences.
If you have a presentation or a poster that you have presented that will not be submitted later
to a journal you could place it here for people to access. When an item is uploaded to precedings it
receives a digital object identiﬁer (DOI) which can be used to cite the material later.
Connotea is the tool that I am responsible for. It is a citation and bookmark management service.
The best thing to do is to probably just create an account and start using the site, if you are
in ﬁnding out how it works and play with it a bit, it's pretty easy to use, and I'll give you a quick
demo of how it works
Social bookmarking sites are often write only -> By making collections public and tagged you
provide a resource -> For example I read what my boss is tagging
This is kind of an orthogonal use case for these kinds of services, even before we add in more
for example I'll show you later a tool that joins folksonomies with ontologies, (entity decsriber)
In a social environment you create a collection that might represent your interest, by adding some
intelligence to the back end we hope that we can aid the presence of serendipity by
highlighting related things like related tags, users, and related content.
It might also provide a means to keep track of the scientiﬁc conversation around a topic, outside of
citations in the literature, for example by tying blog comments to doi's of papers.
The Chemical Blogspace can read tags in connotea for items that have a speciﬁc chemical tag. This
way we can begin to automate the aggregation of social information around scientiﬁc pieces of
Mirko Gontek at the university of Colonge
information visualisation of links in connotea
These social links can create networks of information on top of the basic
This is what we want to use to start building collaborative intelligence into
This is the graph of items for one user in connotea.
There are also other tools out there that are doing the same kind of thing, but I’m partial.
I hope that you take away from this talk that tools are beginning to emerge. We are at the very early
days in their development
and than can only get more powerful as the expertise in developing these kinds of sites becomes
A frequent complaint that I get is that people have no time to investigate new ways of working,
however if you look at how you manage your information now a small investment in, for example,
posting your talks to a community site, or having all of your bookmarks in one location, could begin
to make a difference.