Yahoo! Shines at KDD Conference
Yahoo! was in the spotlight at the 13th annual ACM SIGKDD, the premier international conference on knowledge discovery and data mining, held August 12th-15th in San Jose.
Not only did Yahoo! present a number of groundbreaking papers, panels and workshops, it also took home two of the most prestigious prizes from the conference: the Innovation Award and the Best Research Paper Award.
One of the biggest newsmakers at this year’s event was Yahoo!’s Usama Fayyad, who was honored with the 2007 Innovation Award. This is the highest prize in the fields of data mining and knowledge discovery. It is awarded annually to the individual who has most influenced research and development in data mining and knowledge discovery, and who has driven significant innovations in these areas.
In particular, Fayyad was honored for his work in machine learning and data mining algorithms that scale to large commercial database systems, and for the development of fundamental applications in mining massive science data sets that have lead to new scientific discoveries.
"This award highlights the importance of technical innovation in real applications and in building analysis systems at a massive scale, and I’m honored to be recognized," said Fayyad. "At Yahoo!, advancements in data mining and knowledge discovery are making it possible for us to better understand how our customers use our products and what the future of the Internet experience will be."
Fayyad was not only a major award winner at this year’s KDD conference, he was also a featured speaker. His gave an invited talk to approximately 900 attendees called "From Mining the Web to Inventing the New Sciences Underlying the Internet." The talk covered some of the biggest challenges currently facing data mining. Additionally, he spoke about the larger issues surrounding the Internet, including community, personalization, and the new microeconomics of the web—topics near and dear to Yahoo! Research.
The other big Yahoo! Research winners at KDD 2007 were Deepak Agarwal and Srujuna Merugu, who captured the Best Research Paper Award for their work on "Predictive Discrete Latent Factor Models For Large Scale Dyadic Data." In the paper, the authors proposed a novel statistical method to predict large scale dyadic response variables in the presence of covariate information.
All told, more than a dozen scientists from Yahoo! attended KDD 2007. Pavel Berkhin, who served as general chair of this year’s KDD conference, spoke on the main panel, "Mining at the Crossroads: Successes, Failures and Learning from Them."
Accepted research track papers from Yahoo! Research included:
- Efficient and Effective Explanation of Change in Hierarchical Summaries by Deepak Agarwal, Dhiman Barman, Dimitrios Gunopulos, Flip Korn, Divesh Srivastava, and Neal Young
- Estimating Rates of Rare Events at Multiple Resolutions by Deepak Agarwal, Andrei Broder , Deepayan Chakrabarti , Dejan Diklic , Vanja Josifovski , and Mayssam Sayyadian
- Extracting Semantic Relations from Query Logs by Ricardo Baeza-Yates and Alessandro Tiberi
- Feature Selection Methods for Text Classification by Anirban Dasgupta, Petros Drineas, Boulos Harb, Vanja Josifovski , and Michael Mahoney
Also worthy of note, Vanja Josifovski and Mohammed Mahdian were on the program committee for the First International Workshop on Data Mining and Audience Intelligence for Advertising held in conjunction with KDD 2007. This was the very first workshop to focus on data mining techniques for online advertising. Deepayan Chakrabarti and Srujana Merugu served as local arrangements co-chairs.
All in all, KDD 2007 was a spectacular showcase for Yahoo! Research—and a telling reflection of the outstanding work being accomplished by everyone.
