The Man in the Driver's Seat at Yahoo! Research Barcelona

When Yahoo! Research wanted to extend its reach into Europe, to draw in top talent from the Continent and facilitate collaboration with researchers there, they were fortunate to grab renowned Web retrieval and mining expert Ricardo Baeza-Yates to lead the new Yahoo! Research laboratory in Barcelona, Spain.

Baeza-Yates, co-author of one of the most widely-used books on information retrieval, came to Barcelona in 2005 to join the faculty of Pompeu Fabra University. He was drawn there by a government program, called Catalan Institution of Research and Advanced Studies (ICREA), designed to foster world-class research in Catalonia, of which Barcelona is capital.

"ICREA brought in a lot of good researchers from all over the world in all fields, and that has allowed us to find interesting people here to work with," Baeza-Yates says.

Baeza-Yates is maintaining his ties with the Pompeu Fabra University, as well as the University of Chile, where he founded the Center for Web Research and directed it until 2005. Chile has recognized him for his work by making him a member of the Chilean Academy of Sciences—the first computer scientist to be inducted. Baeza-Yates will also head another new Yahoo! Research laboratory in Santiago, Chile.

The Barcelona branch of Yahoo! Research will be run in conjunction with the Barcelona Media Innovation Center, a non-profit institution with support from local industry and the Catalan government, and ties to Pompeu Fabra University and other local universities.

The new lab is now forging collaborations with faculty and students at universities not only in Catalonia but around Europe. They plan to bring in students as interns as well as mentor Ph.D. students working on theses in the laboratory.

The work at Yahoo! Research Europe in Barcelona will focus primarily on the interrelated areas of Web mining and search.

Mining includes digging through the content of pages, studying the structure of the network of linked webpages, and sifting through data recording how people use search engines, Baeza-Yates says.

In 2000, Baeza-Yates created a search engine, called TodoCL ("All of Chile") that focuses on Chilean websites. It draws about 2,000,000 users each month—a rich source of data for Baeza-Yates on how people use a search engine.

"But at Yahoo!, I'll have access to far more data," Baeza-Yates says. "You go from gigabytes to terabytes of data."

In the area of Web search, research is not just about developing better algorithms, mathematical rules that decide how to rank search results, Baeza-Yates explains. Improving search also involves improving the speed of results, and how people interact with the search engine—such as the ability of the search engine to make recommendations for pages that are related to, but don't contain, a user's search terms.

The first to join Baeza-Yates at the Barcelona site is Hugo Zaragosa, who was previously at Microsoft Research in Cambridge, England. His research there was a continuation of his doctoral work on probabilistic models of search, and he collaborated with several Microsoft development teams, including MSNSearch and Office. At Yahoo, he will work on search and natural language processing—that is, giving search engines some measure of artificial intelligence so they can understand questions written the way one person speaks to another.