Computer-readable recording medium storing explanatory program, explanatory method, and information processing apparatus

Patent metadata

Field	Value
Publication number	US-2023133868-A1
Application number	US-202217945102-A
Country	US
Kind code	A1
Filing date	Sep 15, 2022
Priority date	Nov 4, 2021
Publication date	May 4, 2023
Grant date	—

Abstract

A recording medium storing an explanatory program for causing a computer to execute an explanatory process. The process includes: generating a plurality of pieces of data based on first data; calculating a ratio of output results, among a plurality of results output in a case that each of the plurality of pieces of data is input to a machine learning model, different from first results output in a case that the first data is input to the machine learning model; generating a linear model based on the plurality of pieces of data and the plurality of results in a case that the calculated ratio satisfies a criterion; and outputting explanatory information with respect to the first results based on the linear model.

First claim

What is claimed is:

1 . A non-transitory computer-readable recording medium storing an explanatory program for causing a computer to execute a process, the process comprising:
generating a plurality of pieces of data based on first data; calculating a ratio of output results, among a plurality of results output in a case that each of the plurality of pieces of data is input to a machine learning model, different from first results output in a case that the first data is input to the machine learning model; generating a linear model based on the plurality of pieces of data and the plurality of results in a case that the calculated ratio satisfies a criterion; and outputting explanatory information with respect to the first results based on the linear model.

2 . The non-transitory computer-readable recording medium according to claim 1 ,
wherein the first data is graph data indicating a graph structure including a plurality of nodes and edges that couple the nodes to each other, and the generating of the plurality of pieces of data includes generating the plurality of pieces of data that satisfies a condition of a designated graph structure based on the first data.

3 . The non-transitory computer-readable recording medium according to claim 1 , the process further comprising:
generating another plurality of pieces of data based on the first data in a case that the ratio does not satisfy the criterion; calculating another ratio of results, among another plurality of results output in a case that each of the another plurality of pieces of data is input to the machine learning model, different from the first results; generating another linear model based on the another plurality of pieces of data and the another plurality of results in a case that the another ratio satisfies the criterion; and outputting another piece of explanatory information with respect to the first results based on the another linear model.

4 . The non-transitory computer-readable recording medium according to claim 1 , the process further comprising:
determining whether to retrain the machine learning model in a case that the ratio does not satisfy a criterion.

5 . The non-transitory computer-readable recording medium according to claim 1 ,
wherein the criterion is such a criterion that the ratio is 60 to 80 percent.

6 . An explanatory method performed by a computer, the method comprising:
generating a plurality of pieces of data based on first data; calculating a ratio of output results, among a plurality of results output in a case that each of the plurality of pieces of data is input to a machine learning model, different from first results output in a case that the first data is input to the machine learning model; generating a linear model based on the plurality of pieces of data and the plurality of results in a case that the calculated ratio satisfies a criterion; and outputting explanatory information with respect to the first results based on the linear model.

7 . The explanatory method according to claim 6 ,
wherein the first data is graph data indicating a graph structure including a plurality of nodes and edges that couple the nodes to each other, and the generating of the plurality of pieces of data includes generating the plurality of pieces of data that satisfies a condition of a designated graph structure based on the first data.

8 . The explanatory method according to claim 6 , the method further comprising:
generating another plurality of pieces of data based on the first data in a case that the ratio does not satisfy the criterion; calculating another ratio of results, among another plurality of results output in a case that each of the another plurality of pieces of data is input to the machine learning model, different from the first results; generating another linear model based on the another plurality of pieces of data and the another plurality of results in a case that the another ratio satisfies the criterion; and outputting another piece of explanatory information with respect to the first results based on the another linear model.

9 . The explanatory method according to claim 6 , the method further comprising:
determining whether to retrain the machine learning model in a case that the ratio does not satisfy a criterion.

10 . The explanatory method according to claim 6 ,
wherein the criterion is such a criterion that the ratio is 60 to 80 percent.

11 . An information processing apparatus comprising:
a memory, and a processor coupled to he memory and configured to perform a process including: generating a plurality of pieces of data based on first data; calculating a ratio of output results, among a plurality of results output in a case that each of the plurality of pieces of data is input to a machine learning model, different from first results output in a case that the first data is input to the machine learning model; generating a linear model based on the plurality of pieces of data and the plurality of results in a case that the calculated ratio satisfies a criterion; and outputting explanatory information with respect to the first results based on the linear model.

12 . The information processing apparatus according to claim 11 ,
wherein the first data is graph data indicating a graph structure including a plurality of nodes and edges that couple the nodes to each other, and the generating of the plurality of pieces of data includes generating the plurality of pieces of data that satisfies a condition of a designated graph structure based on the first data.

13 . The information processing apparatus according to claim 11 , the process further including:
generating another plurality of pieces of data based on the first data in a case that the ratio does not satisfy the criterion; calculating another ratio of results, among another plurality of results output in a case that each of the another plurality of pieces of data is input to the machine learning model, different from the first results; generating another linear model based on the another plurality of pieces of data and the another plurality of results in a case that the another ratio satisfies the criterion; and outputting another piece of explanatory information with respect to the first results based on the another linear model.

14 . The information processing apparatus according to claim 11 , the process further including:
determining whether to retrain the machine learning model in a case that the ratio does not satisfy a criterion.

15 . The information processing apparatus according to claim 11 , wherein the criterion is such a criterion that the ratio is 60 to 80 percent.

Assignees

• Fujitsu Ltd

Inventors

• Todoriki Masaru
• Maruhashi Koji

Classifications

• G06N5/045 Primary

  Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence · CPC title

• G06F18/2451

  linear, e.g. hyperplane · CPC title

• G06N20/00 Primary

  Machine learning · CPC title

• G06F18/2323

  based on graph theory, e.g. minimum spanning trees [MST] or graph cuts · CPC title

• G06K9/6224

  Physics · mapped topic