Dynamic updating of digital data

Patent metadata

Field	Value
Publication number	US-12026148-B2
Application number	US-202117210835-A
Country	US
Kind code	B2
Filing date	Mar 24, 2021
Priority date	Mar 24, 2021
Publication date	Jul 2, 2024
Grant date	Jul 2, 2024

Abstract

Embodiments of the present invention provide a computer system a computer program product, and a method that comprises converting the retrieved data to a uniform syntax for data assessment; performing a query on a plurality of external data sources for additional information associated with the converted data; analyzing a plurality of indicative markers associated with the retrieved data and the additional information; generating a plurality of machine learning models associated with the converted data based on the analysis of each indicative markers within the plurality of indicative markers; dynamically selecting at least one generated machine learning model within the plurality of generated machine learning models associated with the retrieved data based on an analysis of the plurality of indicative markers associated with the retrieved data and the additional information; and automatically verifying an accuracy value associated with the at least one selected generated machine learning model.

First claim

What is claimed is:

1. A computer implemented method comprising:
identifying data retrieved from a server computing device and scanning an external data score for additional information associated with retrieved data;
defining additional information, associated with the retrieved data, wherein the additional information comprises: an origin, authenticity, and an update associated with the retrieved data;
defining a plurality of parameters for dynamically selecting the data from the server computing device, wherein the plurality of parameters comprise: domain, topic, an author, credibility score of the author, area of specialization of the author, and past acceptance information;
converting the retrieved data to a uniform syntax for data assessment by transforming the data to a common dominator using a data translation algorithm, wherein the uniform syntax is able to facilitate uniform comparison, aggregation, and separation of data sets within the retrieved data, and wherein the uniform syntax is an arrangement of words, numbers, symbols, or code to create a formed sentence in a language;
performing a query on a plurality of external data sources for additional information associated with the converted data;
analyzing a plurality of indicative markers associated with the retrieved data and the additional information;
generating a plurality of machine learning models associated with the converted data based on the analysis of each indicative markers within the plurality of indicative markers;
dynamically selecting at least one generated machine learning model within a plurality of generated machine learning models associated with the retrieved data based on an analysis of the plurality of indicative markers associated with the retrieved data and the additional information;
automatically verifying an accuracy value associated with the at least one selected generated machine learning model, wherein the accuracy value is calculated value of correctness associated with the selected generated machine learning model; and
generating a user recommendation for the server computing device associated with the selected machine learning model, wherein the user recommendation is a plurality of ameliorative actions.

2. The computer-implemented method of claim 1 , wherein periodically retrieving data comprises:
identifying digital data stored on the external data sources using a plurality of sensor devices and an artificial intelligence algorithm; and
retrieving the identified digital data from the external data sources.

3. The computer-implemented method of claim 1 , wherein converting the retrieved data comprises encoding the retrieved data from one syntax domain to a different syntax domain.

4. The computer-implemented method of claim 1 , wherein analyzing the plurality of indicative markers associated with the retrieved data and the additional information comprises determining a positive match percentage based on a calculated comparison between at least one data set within the retrieved data and the additional information.

5. The computer-implemented method of claim 1 , wherein dynamically selecting the at least one generated machine learning model comprises:
determining a range of accuracy associated with each machine learning model within a plurality of machine learning models associated with the retrieved data;
assigning an accuracy value for each indicative marker associated with each machine learning model based on a calculated positive match percentage; and
ranking the plurality of machine learning models associated the retrieved data based on the assigned accuracy values.

6. The computer-implemented method of claim 1 , wherein automatically verifying an accuracy value associated with the selected generated machine learning model comprises:
determining when the selected machine learning model associated with the retrieved data was last updated;
comparing the selected machine learning module to previous machine learning model and the additional information retrieved from the external data sources; and
dynamically updating the selected machine learning model with correct information in response to determining that the selected machine learning model contains outdated or incorrect information.

7. The computer-implemented method of claim 6 , wherein dynamically updating the selected machine learning model comprises incorporating the additional information from the external data sources into the selected machine learning model using a machine learning algorithm.

8. A computer program product comprising:
one or more computer readable storage media and program instructions stored on the one or more computer readable storage media, the program instructions comprising:
program instructions to identify data retrieved from a server computing device and scanning an external data score for additional information associated with retrieved data;
program instructions to identify define additional information, associated with the retrieved data, wherein the additional information comprises: an origin, authenticity, and an update associated with the retrieved data;
program instructions to define a plurality of parameters for dynamically selecting the data from the server computing device, wherein the plurality of parameters comprise: domain, topic, an author, credibility score of the author, area of specialization of the author, and past acceptance information;
program instructions to convert the retrieved data to a uniform syntax for data assessment by transforming the data to a common dominator using a data translation algorithm, wherein the uniform syntax is able to facilitate uniform comparison, aggregation, and separation of data sets within the retrieved data, and wherein the uniform syntax is an arrangement of words, numbers, symbols, or code to create a formed sentence in a language;
program instructions to perform a query on a plurality of external data sources for additional information associated with the converted data;
program instructions to analyze a plurality of indicative markers associated with the retrieved data and the additional information;
program instructions to generate a plurality of machine learning models associated with the converted data based on an analysis of each indicative markers within the plurality of indicative markers;
program instructions to dynamically select at least one generated machine learning model within the plurality of generated machine learning models associated with the retrieved data based on the analysis of the plurality of indicative markers associated with the retrieved data and the additional information;
program instructions to automatically verify an accuracy value associated with the at least one selected generated machine learning model, wherein the accuracy value is calculated value of correctness associated with the selected generated machine learning model; and
program instructions to generate a user recommendation for the server computing device associated with the selected machine learning model, wherein the user recommendation is a plurality of ameliorative actions.

9. The computer program product of claim 8 , wherein the program instructions to periodically retrieve data comprise:
program instructions to identify digital data stored on the external data sources using a plurality of sensor devices and an artificial intelligence algorithm; and
program instructions to retrieve the identified digital data from the external data sources.

10. The computer program product of claim 8 , wherein the program instructions to convert the retrieved data comprise program instructions to encode the retrieved data from one syntax domain to a different synta

Assignees

• IBM

Inventors

• Ananthapur Bache Vijay Kumar
• Rangarajan Arvind
• Jayarathanasamy Pradeep Raj
• Thangaraj Srithar Rajan

Classifications

• G06N20/00

  Machine learning · CPC title

• G06F16/2379 Primary

  Updates performed during online database operations; commit processing · CPC title