Information integration

Chris Baker (second from the left) is one of a core group of professors at UNB researching in Semantic Technology, these include Weichang Du, Harold Boley, Yevgen Biletsky, and Virenda Bhavshar. Josh O'Kane photo.Imagine a search engine that didn’t just use what you typed— it understood it, and gave you the results you needed. With semantic searching, that’s possible.

This is the research focus of Chris Baker, one of the world’s leading thinkers in semantic web technology — formal standards and technologies that allow machines to understand the meaning, or “semantics,” of information on the web.

Baker is adapting the technology in a sweeping number of areas, from life science to telecom and beyond. As the Innovatia Research Chair at the University of New Brunswick’s Saint John campus, he’s bringing semantic web technology to the contact centre workers in the local telecommunications industry.

Rather than just spitting out results, semantic search technologies can recognize the meaning of the typed information; it uses metadata, or data about data, and rich hierarchies of information to break down a query and pull up the relevant information you need.

“If you call up a customer service centre on the phone, you’ll be talking to an agent who’s pulling up PDF files and searching those files,” Baker explains. “By taking all that documentation and tagging it with semantic groups, we make it easier to locate the right information. We’re looking at a 50-per cent increase in their ability to find answers.”

Partnering with industry veterans Innovatia, explains Baker, means the development of semantic solutions stays in the practical realm, rather than the theoretical. On top of telecommunications, he already sees applications in the nursing and real estate industries.

“Having a two-way interaction between the company and their search really speeds up the whole innovation process.”

Saving time with semantics

Baker holds a PhD in microbiology, and spent many years doing research in molecular biology and protein engineering before getting involved with a Montreal startup in the late 1990s doing drug discovery. At the dawn of genome sequencing, Baker found himself working on genome scanning, gene discovery and protein domain annotation with the startup.

“The system helped identify new biological pathways for drug synthesis,” he says.

He was eventually recruited as a “domain expert” — the expert, in this case, in fungal genomics — for a computer science project at Concordia University in the early 2000s.  It was one of the first successful projects to adapt semantic technology to life sciences. “That’s where I cut my teeth with semantics,” Baker jokes. “We were pioneers!”

After that, he took on the challenge of communicating the value of semantic technology to life scientists. Together with Kei Cheung of Yale University, he spent a year co-editing the book Semantic Web: Revolutionizing Knowledge Discovery in th Life Sciences, which was published by Springer in 2007. This gave him a birds-eye view of the synergies of semantics with existing technologies and their suitability in multiple life science scenarios.

By combining semantic metadata and text mining into discovery platforms, scientists can more easily search through research papers.

“Five hundred search results can become 50 sentences that are very precise answers” explains Baker.

It’s a time saver that works.

Semantics can also help unify naming systems that aren’t well-standardized, like that of lipids. Lipids, such as cholesterol and omega fatty acids, are often are named according to biological or other features, as opposed to their chemical structures — but with semantic technology, structural information about all lipids can be integrated as rules and used to compare with new lipid structures to determine correct structural names.

Determining a lipid’s systematic name on one’s own might takeup to a year to do, delaying publication or other potential discoveries; with semantic web technology, scientists can spend their time applying their knowledge rather than just integrating data or browsing for answers.

Commercial appeal

Baker, a Londoner from the U.K., has spent close to 10 years working on semantic web technology across the globe, from Montreal to Singapore and back. UNB was the next natural step, he explains, because it gave him the chance to work in direct partnership with industry to create and evaluate practical adaptations of the technology in the role of Innovatia Research Chair.

“One of my experiences in trying to get new technology adopted was that there’s often a disconnect between scientists and people in the business world trying to deploy solutions. They talk in different languages,” he says. “I saw this as a good setup in which to develop technology to make sure it gets into a commercial context.”

Working in that commercial context means working to make the technology more accessible to the user.

“We need intelligent systems that can communicate to us when we search. I don’t want to spend 10 minutes searching through the knowledgebase to find that what I want is not there.”

With the right efforts, Baker believes that semantic web technology will soon become a new standard integrated into search tools and that it will provide better query composition options, inference and classification as well as work toward meaningful data integration.

Bakers work at UNB is funded by New Brunswick Innovation Foundation, the Atlantic Canada Opportunities Agency, the Natural Sciences and Engineering Research Council of Canada Discovery Grants Program, Canada’s Advanced Research and Innovation Network (CANARIE) and the Québec New Brunswick University Co-operation in Advanced Education and Research program, Collaborative Health Research Projects Program from Canadian Institutes of Health Research (CIHR) and NSERC.

Contributed by Josh O’Kane. This story was made possible thanks to the financial support of the UNB Associated Alumni

Tagged as: , , , , ,

Comments are closed.