Michaela Götz

Title: Michaela Götz: Innovator in Active Learning Record Matching

Introduction

Michaela Götz is a prominent inventor based in Ithaca, NY (US). She has made significant contributions to the field of record matching through her innovative patent. Her work focuses on enhancing the efficiency and accuracy of identifying duplicate records, which is crucial in various data management applications.

Latest Patents

Michaela holds a patent titled "Active learning of record matching packages." This patent describes an active learning record matching system and method designed to produce a record matching package that identifies pairs of duplicate records. The system allows for a precision threshold to be specified, generating a learned record matching package with precision exceeding this threshold and recall close to the optimal level. The method employs a blocking technique to limit the scope of record matching packages considered, enabling scalability to large datasets. Additionally, the learning method evaluates multiple record matching packages, estimating their precision and recall, and identifies the package with maximum recall that meets the specified precision threshold. A human domain expert is involved in labeling a sample of record pairs as matches or non-matches, which is then used to estimate the package's precision.

Career Highlights

Michaela is currently associated with Microsoft Technology Licensing, LLC, where she continues to innovate and contribute to advancements in technology. Her work is instrumental in improving data processing and management systems.

Collaborations

Michaela has collaborated with notable colleagues, including Arvind Arasu and Shriraghav Kaushik, who share her commitment to advancing technology in the field of record matching.

Conclusion

Michaela Götz is a trailblazer in the realm of active learning record matching, with her innovative patent paving the way for more efficient data management solutions. Her contributions are vital in addressing the challenges of duplicate record identification in large datasets.