Cookie policy: This site uses cookies to simplify and improve your usage and experience of this website. Cookies are small text files stored on the device you are using to access this website. For more information on how we use and manage cookies please take a look at our privacy and cookie policies. Your privacy is important to us and our policy is to neither share nor sell your personal information to any external organisation or party; nor to use behavioural analysis for advertising to you.

Research intelligence

Help to make missing links

Elsevier wants to foster web tools that 'improve' science by making the oceans of data online more navigable. Zoë Corbyn reports

New web tools designed to help academics deluged by electronic information digest and interpret research papers are being developed - prompting claims that they will improve not just scientific publishing but will "improve science itself".

Last month, the multinational scientific publisher Elsevier announced the results of its Grand Challenge Competition - a global search for the best prototype tools to help researchers manage the ever-growing fund of online information in the field of life sciences.

"The objective was to generate useful new ideas that could have a widespread impact on scientific publishing in general," explained Anita de Waard, principal researcher in disruptive technologies at the Elsevier Labs in Amsterdam.

"The bread and butter of scientific communication is still the scientific paper ... but there are so many papers in so many fields that it is overwhelming ... We need new technologies to interpret all the data."

The tools fall under the rubric of "semantic web technology". This aims to make web pages - which are designed to be read by people - more understandable to computers so that machines can perform some of the more mechanical duties of finding, sharing and combining data online.

Opportunities abound in the life sciences, where there is already much collaboration between biologists and computer scientists.

"There is not a biologist today who does not rely heavily on computers to solve certain issues for them," Dr de Waard said.

Many researchers, including ones at Elsevier, are already working on "fact-extraction" or "claim-identification" technology. The idea is that computers sift out the hypotheses or claims made in papers and match them with the evidence. This saves time and makes it easier to see where papers agree with or contradict each other.

The work, which is in its early stages, is tricky because there is much in a scientific text that is impossible to model, Dr de Waard said.

Automatic annotation

The winner of Elsevier's competition, which received a cash prize of $35,000 (£22,400), was a tool called Reflect. This automatically annotates scientific terms within papers. It is already available online free of charge for researchers to access (http://reflect.ws).

Reflect is the brainchild of a team at the European Molecular Biology Laboratory (EMBL) in Germany.

The program - which can be installed on Firefox or Internet Explorer web browsers - "tags" items such as genes, proteins or small molecules mentioned in a paper and opens windows that offer a wealth of information about each, sourced from commonly used life-science databases.

The program, which also works for chemicals, is being extended to cover the medical realm.

"The tool is a shortcut," said Reinhard Schneider, the team leader of EMBL's computational biology programme. "Even if you are an expert in your field, you come across protein names or chemicals and you don't know what they do. You might google them or search databases. But with Reflect, you just hit a button and you have the most important information you need there in the pop-up. If you need more information, you can click through to the primary source."

Elsevier is exploring whether the tool could be incorporated into Cell, its leading experimental biology journal, Dr de Waard said.

It is also looking at the potential for commercially developing the eight other semifinalists.

Implicit connections

The competition runner-up, which received $15,000, was Coraal, a tool developed by the National University of Ireland.

It is a prototype for intelligently searching the contents of articles that already spans 11,000 Elsevier papers related to cancer research. Unlike search engines such as Google, it uses the relationships between words to discern knowledge that is not explicit. It, too, is available free for researchers to use (http://coraal.deri.ie:8080/coraal).

Dr de Waard said the competition allowed Elsevier to collaborate with the scientific community in developing innovative ways to get the most out of scientific papers. It was about "Elsevier standing up and saying 'we need you scientists and all your ideas' and we don't think we can do this by ourselves".

The tools would help improve not just scientific publishing, but science itself, she said. "(These tools offer) ways in which you can more easily read articles outside your immediate field or connect with different researchers."

The company plans to hold a conference on the future of research communication in Boston next March.

zoe.corbyn@tsleducation.com.

  • Print
  • Share
  • Save
  • Print
  • Share
  • Save