Location History:
- Princeton, NJ (US) (2012 - 2013)
- San Francisco, CA (US) (2014)
Company Filing History:
Years Active: 2012-2014
Title: David Grangier: Innovator in Data Classification and Semantic Indexing
Introduction
David Grangier is a notable inventor based in Princeton, NJ (US). He has made significant contributions to the fields of data classification and semantic indexing. With a total of 3 patents, Grangier's work focuses on enhancing the accuracy and efficiency of data processing systems.
Latest Patents
Grangier's latest patents include "Feature set embedding for incomplete data" and "Supervised semantic indexing and its extensions." The first patent discloses methods and systems for classifying incomplete data. This method generates pairs of features and values based on measurements from the incomplete data. A transformation function is then applied to these pairs to create a set of vectors in an embedding space. A hardware processor subsequently applies a prediction function to these vectors, generating confidence assessments for various classes of data. The second patent outlines a system for determining the similarity between a document and a query. It utilizes both frequently and infrequently used dictionaries to correlate words and grams, ultimately determining a similarity score between weight vectors of a query and documents in a corpus.
Career Highlights
David Grangier is currently employed at NEC Laboratories America, Inc. His work at this company has allowed him to explore innovative solutions in data processing and classification. His patents reflect a deep understanding of the complexities involved in handling incomplete data and semantic relationships.
Collaborations
Grangier has collaborated with notable colleagues, including Bing Bai and Jason Edward Weston. These collaborations have likely contributed to the development of his innovative patents and research.
Conclusion
David Grangier stands out as an influential inventor in the realm of data classification and semantic indexing. His contributions through his patents have the potential to significantly impact how data is processed and understood in various applications.