Triple The Fun in Linked Data Challenge

By SemanticWeb.com on September 10, 2009 5:14 PM

Jennifer Zaino
SemanticWeb.com Contributor

The awards have just been handed out for the 2009 Linking Open Data Triplification Challenge. The awards are designed to raise awareness and showcase best practices around revealing and exposing existing structured representations, which are already backing most Web sites.

First prize in the competition this year went to the Linking Open Drug Data initiative, spearheaded by Anja Jentzsch, Jun Zhao, Oktie Hassanzadeh, Kei-Hoi Cheung, Matthias Samwald, and Bo Andersson. The initiative succeeded in bringing a vast amount of heterogeneous information to the semantic data web, says Sören Auer, senior scientist and head of the research group Agile Knowledge Engineering and Semantic Web at the Institute for Applied Computer Science at the University of Leipzig. Auer initiated the first Triplification challenge last year and was a member of this year’s organizing committee.

The Drug Bank, Daily Med, Sider, RDF-TCM (Traditional Chinese Medicine), Linked CT, and Diseasome are the data sets published and interlinked by the project so far. As of August the data sets consisted of more than 8 million RDF triples, which are interlinked by more than 370,000 RDF links. The project hopes to demonstrate how researchers in life science, as well as physicians and patients, can take advantage of connected data sets – discovering links between traditional Chinese and western medicine, for instance, that can connect potential side effects from using an herb to side effect information reported in clinical trials of western drugs that share similar ingredients.

The winner of the second prize developed a new and innovative approach for exposing structured Linked Data from information stored in traditional file system hierarchies, says Auer of Bernhard Schandl’s TripFS. The third prize was given to a project titled Standardized Multilingual Language Resources for the Web of Data. It mapped a large corpus of multilingual language resources to the semantic data web, thus improving information exchange for linguists, Auer says.

An honorable mention

“Since the quality of the submissions was overall very high this year, the program committee decided to award an honorable mention to SensorMasher, which aims at connecting a network of small sensors with the data web, and to PoolParty, which is a very easy to use user interface for creating taxonomic linked data sources,” he notes.

Fifteen submissions in total were received. Approximately two-thirds were submitted from academia and the other third from industry sources.

“With Open Drug Data and PoolParty two finalists are demonstrating the strong relevance of linked data for the enterprise,” Auer comments. “The integration and use of linked data source in corporate environments … have a huge potential. We hope to see more submissions next year which demonstrate how Linked Data technologies can be used in corporate environments.”

A particular focus in 2010, he thinks, will revolve around how linked data technologies can be used for integrating information in corporate data intranets.

A notable trend among this year’s submissions was that they covered a variety of different domains. “Compared to last year’s challenge (where submissions were mainly domain- agnostic), a number of submissions this year demonstrated the benefits of linked data in very concrete domains,” he says, including Linked Open Drug Data, Sensormasher, and the Language Resources entries.

It was harder to choose the third prize winner and honorable mentions this year, given the number of strong submissions the committee saw, Auer says. He believes that that shows that linked data is becoming a more and more mature technique and the community of researchers, developers and users is steadily growing.

Comments