Elsevier Challenges Contestants to Solve Information Overload Dilemmas


      Bookmark and Share

Article ImageElsevier, a global healthcare and scientific publisher, announced the four finalists chosen in the first Elsevier Grand Challenge on Jan. 6. The competition invited researchers to prototype tools dealing with the ever-increasing amount of online life sciences information. With cash prizes of $35,000 for first place and $15,000 for second, the winners have quite a bit at stake. Finalists will move on to present their solutions for the judges via webinar at the Experimental Biology conference in April.

"About a year and a half ago, I was discussing the potential of providing access to Elsevier content to help researchers resolve the challenge of accessing academic information … at the Information Sciences Institute of the University of Southern California. This idea developed into the creation of an open innovation competition, focused on improving the way scientific information is communicated and used," says Anita de Waard, co-organizer of the challenge and researcher, disruptive technologies, Elsevier Labs.

It has been a long road getting the Grand Challenge from an idea to the final stages of choosing a winner. "We launched the competition in June 2008. From the 72 entries we received (in abstract form), we selected 10 semi-finalists. One withdrew, which left us with nine semi-finalists, who worked on our data for 4 months and further developed their ideas," says de Waard. The groups presented their projects to each other and the judges at a December 2008 meeting on MIT’s campus in Boston. From there, the four finalists were picked.

The projects are based on a range of technology solutions focused on increasing the accessibility of life sciences information—doing everything from highlighting genes, proteins, and small molecule names in any webpage to extracting metadata and knowledge from an article and then piecing it together with emergent content and displaying it in a multiple-perspective search/browse interface. The final teams all hail from different countries and walks of life. According to de Waard, "Competitors were mostly academics, but coming from very varied fields, ranging from computational linguistics and machine learning to semantic web technologies and biology. We also had some entries from industry, although they didn’t get through the semi-finals."

Seán I. O’Donoghue, Lars Jensen, Heiki Horn, Evangelos Pafilis, Michael Kuhn, Nigel P. Brown, and Reinhard Schneider, representing Germany, put forth a project called "Reflect: Automated Annotation of Scientific Terms." From Ireland, Vit Novacek, Tudor Groza, Ioana Hulpus, and Siegfried Handschuh entered "CORAAL—Dive into Publications, Bathe in the Knowledge." Amr Ahmed, Andrew Arnold, Luis Pedro Coelho, Joshua Kangas, Abdul-Saboor Sheik, Eric Xing, William Cohen, and Robert F. Murphy of the U.S. put together "Structured Literature Image Finder." Last but not least is a team from Australia—Stephen Wan, Cecile Paris, Robert Dale, Michael Muthukrishna, Ilya Anisimoff, and Julien Blondeau—with "Citation Sensitive In-Browser Summarisation of Cited Documents."
Elsevier has three goals for this competition: First, the company would like to see what kind of innovation is "out there, beyond Elsevier’s walls." Next, Elsevier hopes to spread the word that it is interested in these types of proposals. Finally, they would like to build a community focused on these solutions.

The finalists seem to be forging that community already. Novacek, of the CORAAL prototype team says, "It was great to meet the challenge judges, contestants, and organizers in person at the semi-final presentation day at MIT. We had a couple of interesting discussions with other contestants, some of which may turn into future cooperations between our research groups—namely regarding semantic publishing."

Novacek is not the only one making new contacts and possible partnerships through the contest. Dr. Seán O’Donoghue says, "Personally, I made several significant contacts at the meeting—we will probably work with two of the contestants and with Elsevier directly, especially ScienceDirect, on Reflect. And I plan to collaborate with one of the other contestants on another project."

To de Waard’s mind, this community building is one of the best things to come out of the challenge: "It has been interesting watching the interaction between the judges and competitors, and I see the ongoing relationship between these people and Elsevier as one of the most important benefits we could achieve coming out of the competition."

(www.elseviergrandchallenge.com; www.elsevier.com/wps/find/authored_newsitem.cws_home/companynews05_01186)