Impact score for ontology changes

Patent metadata

Field	Value
Publication number	US-2024256591-A1
Application number	US-202318162744-A
Country	US
Kind code	A1
Filing date	Feb 1, 2023
Priority date	Feb 1, 2023
Publication date	Aug 1, 2024
Grant date	—

Abstract

Described are techniques for a re-analysis of assignments of terms to assets. The techniques include detecting a change in a term ontology comprising a plurality of terms, and determining at least one selected from a group consisting of: a domain feature change vector (DFCV) for a domain of the term ontology affected by the change, and a term feature change vector (TFCV) for the term affected by the change. The techniques further include identifying assets for the re-analysis of the assignments of terms, wherein each of the identified assets is associated with an impact score value based on the DFCV and/or the TFCV, and performing the re-analysis of the assignments of terms for the identified assets ordered by the impact score value.

First claim

What is claimed is:

1 . A computer-implemented method for a re-analysis of assignments of terms to assets, comprising:
detecting a change in a term ontology comprising a plurality of terms; determining at least one selected from a group consisting of: a Domain Feature Change Vector (DFCV) for a domain of the term ontology affected by the change, and a Term Feature Change Vector (TFCV) for a term affected by the change; identifying assets for the re-analysis of the assignments of terms, wherein each of the assets is associated with an impact score value based on the DFCV and/or the TFCV; and performing the re-analysis of the assignments of terms for the assets ordered by the impact score value.

2 . The computer-implemented method of claim 1 , wherein identifying the assets for re-analysis further comprises:
determining the impact score value based on the DFCV and/or the TFCV for a plurality of assets out of which the identified assets are selected if an associated impact score value is equal to or greater than a predefined first threshold value.

3 . The computer-implemented method of claim 1 , wherein the re-analysis is performed for the identified assets ordered by the impact score value until an associated impact score value is equal to or less than a predefined second threshold value, or the re-analysis yields no further change in the assignments of the terms.

4 . The computer-implemented method of claim 1 , wherein the re-analysis is performed in an order of decreasing impact score values, and/or wherein the impact score values associated with the assets are indicative of at least one selected from a group consisting of:
a likelihood for a change of the assignments of terms to respective assets, an expected change of an assignment confidence value for the terms assigned to the respective assets, and an expected change of assignment quality value due to a change of the assignments of terms to the respective assets.

5 . The computer-implemented method of claim 1 , wherein the change in the term ontology comprises at least one selected from a group consisting of:
a term being added to the term ontology, a term being removed from the term ontology, a change of a term name of the term in the term ontology, a change of a term description of the term in the term ontology, a change of a term relation between at least two of the terms in the term ontology, a term split of a term in the term ontology, and a term union of at least two of the terms in the term ontology.

6 . The computer-implemented method of claim 1 , wherein each of the assignments of terms comprises at least one selected from a group consisting of:
an assignment confidence value indicative of a confidence that an assigned term matches a respective asset, an indicator of a type of an analysis of the respective asset used to create the assignment, an indicator of a type of the re-analysis of the respective assignment, and an assignment feature vector (AFV), wherein components of the AFV are indicative of a weight value for each feature of the assigned term, wherein the analysis or the re-analysis determines whether or not to assign the term to the respective asset depending on the features weighted according to the respective weight values.

7 . The computer-implemented method of claim 6 , wherein the features of the terms comprise at least one selected from a group consisting of:
a name of the respective term, a description of the respective term, a term relation of the respective term to one or more other terms in the term ontology, an asset relation of the respective term to one or more other assets, a data class of the respective term, a classification of the respective term, and a domain in the term ontology which comprises the respective term.

8 . The computer-implemented method of claim 1 , wherein the impact score value associated with a respective asset is determined based on at least one selected from a group consisting of:
a scalar product of the TFCV and the AFV of each term assigned to the respective one of the assets; and a scalar product of the DFCV and the AFV of each term assigned to the respective one of the assets.

9 . The computer-implemented method of claim 1 , wherein the assets are identified, or the impact score values are determined, based on the TFCV when a change of a single term in the term ontology is detected, or
wherein the assets are identified or the impact score values are determined based on the DFCV when the detected at least one change comprises at least one selected from a group consisting of: changes of multiple terms in the term ontology, a term being added to the term ontology, and a term being removed from the term ontology.

10 . The computer-implemented method of claim 1 , wherein components of the TFCV comprise a predefined or maximum value for each feature of a removed term or for each feature of an added term, or
wherein one or more components of the TFCV are zero when the corresponding one or more features of the term are not affected by the determined at least one change, or wherein the DFCV of a domain of the term ontology is a sum of the TFCVs determined for the terms in the domain.

11 . The computer-implemented method of claim 1 , wherein the assets are physical assets, customer tables, documents, or metadata of source assets.

12 . The computer-implemented method of claim 1 , further comprising:
using a machine learning system or a data classification system for the identification of the assets or for the determination of the impact score values.

13 . The computer-implemented method of claim 12 , wherein the machine learning system is continuously re-trained based on new data of a first or second threshold value for changes of the assignments of the terms.

14 . A re-analysis system for assignments of terms to assets, comprising:
one or more computer readable storage media storing program instructions and one or more processors which, in response to executing the program instructions, are configured to: detect a change in a term ontology comprising a plurality of terms; determine at least one selected from a group consisting of: a Domain Feature Change Vector (DFCV) for a domain of the term ontology affected by the change, and a Term Feature Change Vector (TFCV) for a term affected by the change; identify assets for the re-analysis of the assignments of terms, wherein each of the assets is associated with an impact score value based on the DFCV and/or the TFCV; and perform the re-analysis of the assignments of terms for the assets ordered by the impact score value.

15 . The re-analysis system of claim 14 , wherein the program instructions configured to cause the one or more processors to identify the assets for re-analysis are further configured to cause the one or more processors to:
determine the impact score values based on the DFCV and/or the TFCV for a plurality of assets out of which the identified assets are selected when the associated impact score value is equal to or greater than a predefined first threshold value.

16 . The re-analysis system of claim 14 , wherein the program instructions configured to cause the one or more processors to perform the re-analysis are further configured to cause the one or more processors to:
perform the re-analysis for the assets ordered by the impact score value until the associated impact score value is equal to or less than a predefined second threshold value, or the re-analysis yields no further change in the assignments of the terms.

Assignees

• IBM

Inventors

• Suhre Oliver
• Hampp-Bahnmueller Thomas
• Gerstl Peter
• Saillet Yannick
• Maier Albert
• Baessler Michael

Classifications

• G06F16/35

  Clustering; Classification · CPC title

• G06F16/367 Primary

  Ontology · CPC title