Our mission is to develop effective tools and techniques for enabling a wide variety of knowledge discovery and data mining based applications and solutions. We aim to achieve this by

• Developing scalable and automated machine learning and statistical modeling based data mining techniques for extracting actionable insights from structured and unstructured data sources.
• Embedding mining techniques in middleware platforms.
• Engaging in consulting and services to drive our research agenda for developing novel predictive modeling solutions for data-rich problems in business and industry.goal

Our current actvities include:

• Systems R&D: Developing parallelized implementation of ProbE predictive modeling data mining technologies for database environments
• Solutions R&D: Developing Cross-Channel Optimized Marketing (CCOM) solution for the retail industry Developing data mining based yield mangement solutions for the travel & transportation industry industry Developing data mining based solutions for market intelligence and competitive insights
• Core R&D: Cost Sensitive Learning, Active Learning, Reinforcement Learning, Correlated Equilibria in Graphical Games, Analysis of Margin Classifiers, Relationships between Information Theory and Classification Learning

Recent solutions accomplishments include the Advanced Targeted Marketing for Single Events (ATM-SE) solution for the Retail and Direct Mail industry, and the Underwriting Profitability Analysis (UPA) application for P&C insurance risk management. These solutions are based upon our core data mining research and development vehicle, the ProbE (Probabilistic Estimation) rule based predictive modeling framework.

In the area of unstructured information mining, we have recently been focused upon solutions for web-based automated market intelligence, exploring fast algorithms for document matching, document clustering, and embedding these in solutions for automatic FAQ generation.

Additional accomplishments span the areas of classification rule generation using logic minimization and contextual merit based feature analysis (the RAMP/R-Mini research prototypes) as well as applications research in the areas of automatic text categorization, portfolio management, and manufacturing quality control.

Revised January 23, 2003