Skip to main content

University Collaborations

Uncertain Data (2009)
Peter Haas, University of Florida, and Rice University
Peter worked with Professor Chris Jermaine, formerly of the University of Florida and now at Rice University, to develop a framework for dealing with uncertain data. Their solution came in the form of the Monte Carlo Database (MCDB), a system capable of handling complex and unforeseen types of uncertainty. It was introduced at the SIGMOD Conference in 2008, and continues to be refined.

Social Accessibility (2009)
CoScripter research group and SUNY Stony Brook
We collaborated with SUNY Stony Brook and IBM's own Tokyo Research Labs to study social accessibility, with the goal of improving the usability and performance of CoScripter.

Collaboration Analysis and Visualization (2009)
Jeffrey Pierce and Stanford University
We awarded a faculty grant to Jeff Heer of Stanford University to conduct joint research in the areas of collaboration analysis and visualization.

Parallelizing Query Optimization (2009)
Guy Lohman and Kyungpook National University
Guy worked with Professor Wook-Shin Han of Kyungpook National University to tackle the issue of parallelizing query optimization.

Data Mining (2008)
Jeffrey Nichols, University of Washington, and University of Rochester
We worked with the University of Washington and the University of Rochester to analyze data mining user interactions on the web and build applications around the gathered data.

Raiding Behavior (2009)
Jeffrey Nichols and Indiana University
We worked with Indiana University to study players. raiding behavior in the popular online multiplayer game World of Warcraft.

Detecting Anomalous Events (2008)
Peter Haas and Stanford University
Peter collaborated with Professor Peter Glynn at Stanford University for a number of years, focusing on detecting anomalous events.

Advanced Techniques in Query Processing and Optimization (2006)
Peter Haas, Volker Markl, and Vijayshankar Raman and University of Waterloo and University of Toronto
Peter Haas, Volker Markl, Vijayshanker Raman, and other researchers at Almaden have been collaborating with Prof. Ihab Iyas (U. Waterloo) and Prof. Nick Koudas (U. Toronto) on a range of topics related to advanced techniques in query processing and optimization.

Methods for Incrementally Maintaining Random Samples (2006)
Peter Haas and TU Dresden
Peter Haas has been conducting joint research with Prof. Wolfgang Lehner and his group at TU Dresden on methods for incrementally maintaining random samples of evolving datasets. Members of the InfoSphere project at Almaden have also been working with the TU Dresden group on developing new methods for automatic discovery of complex metadata. All of this work provides key infrastructure for automated information integration and "next generation" BI and data warehousing.

Ensuring Patient Privacy in Healthcare Information Integration (2006)
Alexandre Evfimievski and Tokyo Institute of Technology
This project aims to develop theoretically sound privacy models and cryptographic tools for patient data integration across multiple healthcare institutions. Our focus on the efficiency of these tools would enable healthcare professionals to timely access any needed part of their patients' medical histories while helping to protect the patients' records against unauthorized disclosure. A part of this challenge is to abstract a model for data confidentiality that is flexible enough to express real-life policies and yet rigorous enough to prove privacy guarantees. Increased privacy protection is essential in the ongoing transition of health services to a digital infrastructure.

Web Search (2005)
Avatar research group and University of Michigan
H V Jagadish at University of Michigan. Work on Web Search: Supporting Transactoinal Queries. Work has resulted in a paper in SIGIR 2006. Collaboration through Yunyao Li, a Ph.D. student at Michigan.

Sampling-based Methods for Estimating Number of Distinct Values (2005)
Peter Haas and University of Texax - Austin
Peter Haas has collaborated with Prof. Lynne Stokes at UT Austin on sampling-based methods for estimating the number of distinct values in a data collection. This estimation problem arises in a large number of applications, including database query optimization and automated information integration.

Auditing Compliance with a Hippocratic Database (2004)
Jerry Kiernan and Carnegie Mellon University
We collaborated with Prof. Christos Faloutsos of CMU to invent an auditing framework for determining whether a database system is adhering to its data disclosure policies. In our framework, users formulate audit expressions to specify the sensitive data subject to disclosure review. These expressions are accepted by an audit component, which returns all queries (deemed "suspicious") that accessed the specified data during their execution. This work is published in VLDB 2004.

Datamining Algorithms (2004)
Peter Haas and Northwestern University and Polytechnic University
Peter Haas has worked jointly with Prof. Peter Scheuermann (Northwestern University) and Herve Bronnimann (Polytechnic University) on techniques for speeding up data mining algorithms by applying these algorithms to carefully constructed data synopses.

Sampling-based Estimation of Query Selectivities (2002)
Peter Haas and University of Wisconsin-Madison
Peter Haas has worked with Prof. Jeff Naughton (U. Wisconsin Madison) on methods for sampling-based estimation of query selectivities and for online processing of aggregation queries.

[an error occurred while processing this directive]