News Release

Social networking shortcut to finding medical experts

Peer-Reviewed Publication

BMC (BioMed Central)

It can be difficult for someone outside of a specialist field to identify subject experts and the ever increasing amount of available data can be bewildering. New research, published in BioMed Central's open access journal, Journal of Biomedical Semantics, describes a method of social network analysis, similar to finding friends on Facebook, able to sift through scientific literature and news articles to identify opinion leaders and media experts.

Pharmaceutical companies and public health programs rely on opinion leaders to clarify and condense research into a format understandable by the general public and employees. While there are already computer programs which are able to link scientists and medical doctors to their areas of knowledge, this is the first system that is able to identify the opinion leaders from these subject lists.

A team from Lnx Research developed an extraction engine, which used text mining technology, to produce a network of subject experts, and find people, organisations and locations associated with a specific heath topic from research articles. The three sets of results from the search were combined, and, using simple rules, people with the most 'hits' who were most likely to be the opinion leaders were identified. Social network analysis was then used to generate a ranked web of linked experts, based on the number of times they were mentioned together in news articles.

The team tested their program, using the topic 'obesity', and generated a network of over 16,000 experts with 100,000 links between them. The people at the heart of the network, with the most connections, tended to be the subject experts. In about an hour this 'find a friend' system found experts on obesity with an accuracy of about 90%.

Dr Siddhartha Jonnalagadda who led this research explained, "This amount of data would be impossible to sort through without the help of a computer. However when we analysed the results from the program we found that the top twenty obesity experts had all contributed significantly to the advancement of science in the literature and in the news, many having won awards for their work."

Dr Jonnalagadda continued, "We were also able to identify a group of people who, although they were not publishing scientists, were still experts, most likely involved in public education or in science reporting."

So although the system was designed with the aim of finding scientific experts it was also able to find celebrities, and people with a high media profile, who support a particular cause. This information would be invaluable to scientists, students and people wanting to get straight to the heart of a subject.

###

Notes to Editors

1. Discovering opinion leaders for medical topics using news articles
Siddhartha Jonnalagadda, Ryan Peeler and Philip Topham
Journal of Biomedical Semantics (in press)

Please name the journal in any story you write. If you are writing for the web, please link to the article. All articles are available free of charge, according to BioMed Central's open access policy.

Article citation and URL available on request on the day of publication.

2. Journal of Biomedical Semantics is an open access, peer-reviewed journal that considers articles which addresses issues of semantic enrichment and semantic processing in the biomedical domain. This includes infrastructure for biomedical semantics: and semantic mining, annotation, and analysis.

3. BioMed Central (http://www.biomedcentral.com/) is an STM (Science, Technology and Medicine) publisher which has pioneered the open access publishing model. All peer-reviewed research articles published by BioMed Central are made immediately and freely accessible online, and are licensed to allow redistribution and reuse. BioMed Central is part of Springer Science+Business Media, a leading global publisher in the STM sector.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.