Infosys has been granted a patent for a system, method, and apparatus for data classification in a machine learning system. The patent describes a method that involves generating a data classifier using training data, calibrating the classifier based on parameters, and receiving user input to enhance the classifier’s accuracy. The method aims to achieve a predefined model accuracy threshold. GlobalData’s report on Infosys gives a 360-degree view of the company including its patenting strategy. Buy the report here.
Smarter leaders trust GlobalData
According to GlobalData’s company profile on Infosys, AI for workflow management was a key innovation area identified from patents. Infosys's grant share as of September 2023 was 52%. Grant share is based on the ratio of number of grants to total number of patents.
Data classification in machine learning system with user input
A recently granted patent (Publication Number: US11720649B2) describes a method and system for data classification in a machine learning system. The method involves generating a data classifier for a first dataset using training data. The data classifier is then calibrated based on a set of parameters until its accuracy matches or exceeds a predefined model accuracy threshold value.
The calibration process includes receiving user input that annotates a presented subset of the first dataset, disambiguating the subset of data based on the user input, and generating an enhanced version of the data classifier using the disambiguated subset of data and the training data. The accuracy value of the data classifier is determined by applying the enhanced version to another subset of the first dataset.
The set of parameters used for calibration can include a certainty score, click utilization value, certainty threshold value, model accuracy value, average machine learning time, or average annotation time. These parameters help in determining the number of candidates included in the presented subset of data and adjusting the certainty threshold value for subsequent iterations of the calibration process.
The system described in the patent includes a processor and memory with programmed instructions. The processor generates the data classifier for the first dataset and iteratively calibrates it based on the set of parameters until the accuracy value meets the predefined threshold. The calibration process involves receiving user input, disambiguating the presented subset of data, generating an enhanced version of the data classifier, and determining the accuracy value by applying the enhanced version to another subset of the first dataset.
The patent also covers a non-transitory computer-readable medium that includes instructions for data classification. When executed by a processor, these instructions cause the processor to generate the data classifier, iterate the calibration process based on the set of parameters, receive user input, disambiguate the presented subset of data, generate an enhanced version of the data classifier, and determine the accuracy value.
Overall, this patent presents a method and system for data classification in a machine learning system that involves iterative calibration of a data classifier based on user input and a set of parameters. The calibration process aims to improve the accuracy of the data classifier by disambiguating the presented subset of data and generating an enhanced version of the classifier.
To know more about GlobalData’s detailed insights on Infosys, buy the report here.
Data Insights
From
The gold standard of business intelligence.
Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.
GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.
GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.
