National Science Foundation
HOME | TOPICS | REGION/STATES | RESOURCES | ABOUT PROTOTYPE | SEARCH
STATE HIGHLIGHT
A Machine Learning Approach to Approximate Record Matching
State: New York

ChoiceMaker Technologies, Inc. has brought to healthcare and service communities, innovative and effective record-matching software whose feasibility depends on the ease with which it could be customized. There are many peculiarities and case-specific details that must be accommodated when deploying a record matching/linkage system.

The cost of such customization can easily overwhelm the cost of the technology and inhibit its broad application in the areas where it is most needed. ChoiceMaker has designed and implemented several innovations that simplify and automate the customization tasks, while further enhancing the software's performance and precision. By using the latest in machine learning and applying training data to the record-matching problem,  the company has developed a solution that helps to ensure reliable databases and is compatible with multiple languages.


  Web Policies and Important Links | Privacy | FOIA | Help | SiteMap  
Bridging NSF Science Research, Education, and Innovation, Copyright 2008 TERC.
Funded by NSF #0737174.
Opinions expressed on this site are those of the contributors and not necessarily those of the National Science Foundation.
NSF funding for this project ended in 2008. At this time the site has been archived.