Time: Wednesday, May 9, 2007 (1:30-2:15pm)
Speaker: Tomasz Imielinski (Ask.com)
Title: Hybrid Search: Combining Structured and Unstructured Data into a Search Engine
Future search will be increasingly based on a combination of Web page content and structured data coming from databases and XML data feeds. Structured data is already widely available on the Web today through the so-called deep Web, but it is still poorly integrated into search. Semi-structured data sources such as Wikipedia are also becoming prevalent. This creates new challenges and opportunities. How will answers to queries be presented? Will the ten blue links paradigm be broken? What would be the fusion of relational database query answers and answers provided through text matching?
Dr. Tomasz Imielinski is Vice President of Relevance & Online Systems at Ask.com and a full professor (on leave) in the Department of Computer Science at Rutgers University.
Imielinski has published nearly 100 scientific papers, received numerous patents, and authored two books. In 1991, he co-invented the concept of "association rule mining", which has become one of the key data mining techniques included in every major commercial software database package. His paper on this subject is one of twenty most-cited papers in all computer science literature, and he received the "Most Influential Paper" award in 2002 from the Very Large Databases (VLDB) onference. For his work on mobile computing, he won the "2003 Test of Time" award from ACM SIGMOD (Association for Computing Machinery Special Interest Group on Data Management).
He is a co-founder of Connotate Technologies, one of the leaders in data extraction from the Web.