18606458. CLASSIFYING DATA OBJECTS simplified abstract (Google LLC)

CLASSIFYING DATA OBJECTS

Organization Name

Google LLC

Inventor(s)

Gregory Sean Corrado of San Francisco CA (US)

Tomas Mikolov of Jersey City NJ (US)

Samuel Bengio of Los Altos CA (US)

Yoram Singer of Palo Alto CA (US)

Jonathon Shlens of San Francisco CA (US)

Andrea L. Frome of Oakland CA (US)

Jeffrey Adgate Dean of Palo Alto CA (US)

Mohammad Norouzi of Richmond Hill (CA)

CLASSIFYING DATA OBJECTS - A simplified explanation of the abstract

This abstract first appeared for US patent application 18606458 titled 'CLASSIFYING DATA OBJECTS

The abstract of this patent application describes methods, systems, and apparatus for classifying data objects based on high-dimensional representations of terms and categories.

• Obtaining high-dimensional representations for terms in a vocabulary and classification data for data objects.
• Computing an aggregate high-dimensional representation for a data object using representations of category labels and scores.
• Identifying the closest term in the vocabulary to the aggregate representation as the category label for the data object.

Potential Applications: - Text classification in natural language processing. - Image classification in computer vision. - Recommendation systems in e-commerce.

Problems Solved: - Efficient and accurate classification of data objects. - Handling high-dimensional data representations effectively.

Benefits: - Improved accuracy in categorizing data objects. - Scalability for large datasets. - Enhanced performance in classification tasks.

Commercial Applications: Title: "Advanced Data Classification Technology for Enhanced Recommendation Systems" This technology can be used in e-commerce platforms to improve product recommendations based on user preferences and behavior, leading to increased sales and customer satisfaction.

Questions about the technology: 1. How does this technology compare to traditional classification methods? 2. What are the potential limitations of using high-dimensional representations for data classification?

Original Abstract Submitted

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for classifying data objects. One of the methods includes obtaining data that associates each term in a vocabulary of terms with a respective high-dimensional representation of the term; obtaining classification data for a data object, wherein the classification data includes a respective score for each of a plurality of categories, and wherein each of the categories is associated with a respective category label; computing an aggregate high-dimensional representation for the data object from high-dimensional representations for the category labels associated with the categories and the respective scores; identifying a first term in the vocabulary of terms having a high-dimensional representation that is closest to the aggregate high-dimensional representation; and selecting the first term as a category label for the data object.