Report preparation program and report preparation method

Patent metadata

Field	Value
Publication number	US-10719242-B2
Application number	US-201916291035-A
Country	US
Kind code	B2
Filing date	Mar 4, 2019
Priority date	Mar 9, 2018
Publication date	Jul 21, 2020
Grant date	Jul 21, 2020

Abstract

A computer-readable recording medium storing therein a report preparation program that causes at least one of storages and a computer coupled to the at least one of storages to execute a process for calculating an individual-feature quantity with respect to an input output per second (IOPS) of the at least one of storages for each piece of time-series data included in a time-series data group with respect to the IOPS; statistically processing the calculated individual-feature quantity of the time-series data group; calculating an entire-feature quantity based on the statistically processing; referring to a learning model generated based on at least one of learning time-series data groups and contents of a report for the at least one of learning time-series data groups, the learning model representing a relationship between an entire-feature quantity and contents of the report; and outputting information on contents of the report corresponding the calculated entire-feature quantity.

First claim

What is claimed is:

1. A computer-readable recording medium storing therein a report preparation program that causes at least one of storages and a computer coupled to the at least one of storages to execute a process, the process comprising:
calculating an individual-feature quantity with respect to an input output per second (IOPS) of the at least one of storages for each piece of time-series data included in a time-series data group with respect to the IOPS;
statistically processing the calculated individual-feature quantity of the time-series data group;
calculating an entire-feature quantity based on the statistically processing;
referring to a learning model generated based on at least one of learning time-series data groups and contents of a report for the at least one of learning time-series data groups, the learning model representing a relationship between an entire-feature quantity and contents of the report; and
outputting information on contents of the report corresponding the calculated entire-feature quantity.

2. The computer-readable recording medium storing therein the report preparation program according to claim 1 ,
wherein the report includes a comment on at least one of pieces of time-series data,
wherein the learning model represents, for each type of comment, a relationship between an entire-feature quantity and an index value indicating an appearance frequency of the type of comment in a report, and
wherein the outputting includes outputting an index value corresponding to a calculated entire-feature quantity for each type of comment.

3. The computer-readable recording medium storing therein the report preparation program according to claim 2 , wherein the outputting includes generating and outputting a comment on at least one of pieces of time-series data included in the time-series data group based on an index value corresponding to the calculated entire-feature quantity for each type of comment.

4. The computer-readable recording medium storing therein the report preparation program according to claim 3 , wherein the outputting includes selecting at least one of pieces of time-series data for generating a comment out of the time-series data group based on the calculated individual-feature quantity or a feature quantity for each piece of time-series data included in the time-series series data group, the feature quantity being other than the calculated individual-feature quantity.

5. The computer-readable recording medium storing therein the report preparation program according to claim 1 , wherein the individual-feature quantity represents a magnitude of a data variation of the time-series data group, and wherein the entire-feature quantity represents a minimum value and a maximum value of the individual-feature quantity.

6. The computer-readable recording medium storing therein the report preparation program according to claim 1 , wherein the time-series data group is plural pieces of time-series data having an identical attribute.

7. The computer-readable recording medium storing therein the report preparation program according to claim 6 , wherein the attribute is a time zone in which each piece of data included in the time-series data is acquired or measured.

8. A method of preparing a report, the method executed by a computer comprising:
calculating an individual-feature quantity with respect to an input output per second (IOPS) of a storage for each piece of time-series data included in a time-series data group with respect to the IOPS;
statistically processing the calculated individual-feature quantity to calculate an entire-feature quantity of the time-series data group;
referring to a learning model generated based on at least one of learning time-series data groups and contents of a report for the at least one of learning time-series data groups, the learning model representing a relationship between an entire-feature quantity and contents of the report; and
outputting information on contents of the report corresponding the calculated entire-feature quantity.

9. A machine learning device comprising:
a memory configured to store time series data and a learning model, the learning model representing a relationship between an entire-feature quantity of at least one learning time-series data group and corresponding learning report; and
a processor coupled to the memory and configured to perform acquiring the target time-series data group;
calculating an individual-feature quantity with respect to an input output per second (IOPS) of infrastructure equipment for each piece of time-series data included in the target time-series data group with respect to the IOPS;
statistically processing the calculated individual-feature quantity of the target time-series data group;
calculating an entire-feature quantity based on the statistically processing;
inputting the calculated entire-feature quantity to the learning model to obtain a report for the target time-series data group, the report including at least one comment on target time-series data within the target time-series data group, the at least one comment prepared based on the calculated entire-feature quantity and attention levels assigned to categories of comments within the learning report; and
outputting the obtained report.

10. The machine learning device according to claim 9 , wherein the corresponding learning report of each of the at least one learning time series data group includes a plurality of comments belonging to different comment categories and an attention levels are assigned for the comment categories within the learning reports, the attention levels being index values for determining importance of comments to be included in the obtained report depending on the input calculated entire-feature quantity.

11. The machine learning device according to claim 10 , wherein
the learning model represents, for each category of comment, a relationship between an entire-feature quantity and an index value indicating an appearance frequency of the category of comment in the corresponding learning report, and
wherein the outputting includes outputting an index value corresponding to the calculated entire-feature quantity for each category of comment.

12. The machine learning device according to claim 11 , wherein the outputting includes generating and outputting a comment on at least one of pieces of time-series data included in the time-series data group based on an index value corresponding to the calculated entire-feature quantity for each type of comment.

13. The machine learning device according to claim 12 , wherein the outputting includes selecting at least one piece of the target time-series data to be associated with the at least one comment of the obtained report based on the calculated individual-feature quantity or a feature quantity for each piece of the target time-series data, the feature quantity being other than the calculated individual-feature quantity.

14. The machine learning device according to claim 9 , wherein the individual-feature quantity represents a magnitude of a data variation of the target time-series data group, and wherein the entire-feature quantity represents a minimum value and a maximum value of the individual-feature quantity.

15. The machine learning device according to claim 9 , wherein the at least one learning time-series data group and the target time-series data group include a plurality of pieces of time-series data having an identical attribute.

16. The machine learning device according to claim 15 , wherein the attribute is a time zone in which each p

Assignees

• Fujitsu Ltd

Inventors

• Saito Yuji
• Uchiumi Tetsuya
• Watanabe Yukihiro
• Kondo Reiko
• Suzuki Kazuhiro
• Asaoka Masahiro

Classifications

• G06F3/0613 Primary

  in relation to throughput · CPC title

• G06N20/00 Primary

  Machine learning · CPC title

• G06F18/217

  Validation; Performance evaluation; Active pattern learning techniques · CPC title

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G06F3/0629

  Configuration or reconfiguration of storage systems · CPC title