# US 2023/0409570 A1 · Machine-learning-based, adaptive updating of quantitative data in database system

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0409570 A1
- Document ID: US20230409570A1
- Status: Published application
- Kind: A1
- Application number: 18/454,230
- Filed: 2023-08-23
- Published: 2023-12-21
- Earliest claimed: 2022-02-09
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Ling QIN; Fu Fei XU; Jie ZHANG; Guo Dong WAN; Yi Lei WANG; Wen Jing SHI; Yuan CAO
- Web page: https://patentgrep.com/patent/US20230409570A1

## CPC

- G06F 16/2453: Query optimisation
- G06F 16/2272: {Management thereof}
- G06F 16/25: Integrating or interfacing systems involving database management systems
- G06N 3/08: Learning methods
- G06N 20/00: Machine learning
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Machine-learning-based, adaptive updating of quantitative data in a database system is provided, which includes training one or more machine learning models to facilitate adaptively updating quantitative data in the database system, and ascertaining an update urgency index for updating the quantitative data for one or more data structures of the database system. The update urgency index is representative of an urgency for updating the quantitative data for the data structure(s) and is based, at least in part, on real-time query metrics. The machine learning model(s) is used to adaptively update the quantitative data, where the adaptively updating is based, at least in part, on the ascertained update urgency index. Processing of a database query is optimized in the database system using the adaptively updated quantitative data.

## Claims

### Claim 1

A computer program product for facilitating processing within a computing environment, the computer program product comprising:
one or more computer-readable storage media having program instructions embodied therewith, the program instructions being readable by a processing circuit to cause the processing circuit to perform a method comprising:
training one or more machine learning models to facilitate adaptively updating quantitative data in a database system;
ascertaining an update urgency index for updating quantitative data for one or more data structures of the database system, the update urgency index being representative of an urgency for updating the quantitative data for the data structure(s) and being based, at least in part, on real-time query metrics;
adaptively updating, using the machine learning model(s), the quantitative data, the adaptively updating being based, at least in part, on the ascertained update urgency index; and
optimizing processing of a database query in the database system using the adaptively updated quantitative data.

### Claim 2 (depends on 1)

The computer program product of claim 1, wherein the real-time query metrics comprise a volume and frequency of database queries to the data structure(s) in the database system.

### Claim 3 (depends on 1)

The computer program product of claim 1, wherein the real-time query metrics are based, at least in part, on collecting M top-most database queries to the data structure(s) in order of processor usage for a time interval and estimating therefor a deviation in usage from an estimated processor usage for the M top-most database queries, the estimated deviation being representative of a performance deviation of the M top-most database queries, where M≥1.

### Claim 4 (depends on 1)

The computer program product of claim 1, further comprising predicting, via the machine learning model(s), an execution time of one or more database queries to the data structure(s), wherein the real-time query metrics are based on a difference between the predicted execution time of the one or more database queries to the data structure(s) in comparison to an actual execution time of the one or more database queries to the data structure(s).

### Claim 5 (depends on 4)

The computer program product of claim 4, wherein the difference between the predicted execution time of the one or more database queries to the data structure(s) and the actual execution time of the one or more database queries to the data structure(s) exceeding a threshold predicts poor query performance, and increases the update urgency index to indicate an urgency for updating the quantitative data of the data structure(s) in the database system.

### Claim 6 (depends on 4)

The computer program product of claim 4, wherein the predicting, by the machine learning model(s), the execution time for the one or more database queries uses regression analysis based on time series data.

### Claim 7 (depends on 1)

The computer program product of claim 1, wherein the adaptively updating comprises adaptively updating the quantitative data based on a relative urgency of the ascertained update urgency index, the adaptively updating including currently proceeding with updating of the quantitative data where the ascertained update urgency index indicates updating of the quantitative data is urgent, and delaying proceeding with updating of the quantitative data where the ascertained update urgency index indicates that the updating is not urgent.

### Claim 8 (depends on 1)

The computer program product of claim 1, further comprising identifying a list of data structures for which the update urgency index is to be ascertained, the one or more data structures being in the list of data structures, the list of data structures including any data structures with major data changes since a last updating of quantitative data, any data structures with no quantitative data, and any data structures with stale quantitative data.

### Claim 9 (depends on 8)

The computer program product of claim 8, further comprising predictively identifying, by the machine learning model(s), a data structure in the database system with degrading query execution performance, wherein the list of data structures further includes the predictively identified data structure with degrading query execution performance.

### Claim 10

A computer system for facilitating processing within a computing environment, the computer system comprising:
a memory; and
a processing circuit in communication with the memory, wherein the computer system is configured to perform a method, the method comprising:
training one or more machine learning models to facilitate adaptively updating quantitative data in a database system;
ascertaining an update urgency index for updating quantitative data for one or more data structures of the database system, the update urgency index being representative of an urgency for updating the quantitative data for the data structure(s) and being based, at least in part, on real-time query metrics;
adaptively updating, using the machine learning model(s), the quantitative data, the adaptively updating being based, at least in part, on the ascertained update urgency index; and
optimizing processing of a database query in the database system using, at least in part, the adaptively updated quantitative data.

### Claim 11 (depends on 10)

The computer system of claim 10, wherein the real-time query metrics comprise a volume and frequency of database queries to the data structure(s) in the database system.

### Claim 12 (depends on 10)

The computer system of claim 10, wherein the real-time query metrics are based, at least in part, on collecting M top-most database queries to the data structure(s) in order of processor usage for a time interval and estimating therefor a deviation in usage from an estimated processor usage for the M top-most database queries, the estimated deviation being representative of a performance deviation of the M top-most database queries, where M≥1.

### Claim 13 (depends on 10)

The computer system of claim 10, further comprising predicting, via the machine learning model(s), an execution time of one or more database queries to the data structure(s), wherein the real-time query metrics are based on a difference between the predicted execution time of the one or more database queries to the data structure(s) in comparison to an actual execution time of the one or more database queries to the data structure(s).

### Claim 14 (depends on 13)

The computer system of claim 13, wherein the difference between the predicted execution time of the one or more database queries to the data structure(s) and the actual execution time of the one or more database queries to the data structure(s) exceeding a threshold predicts poor query performance, and increases the update urgency index to indicate an urgency for updating the quantitative data of the data structure(s) in the database system.

### Claim 15 (depends on 13)

The computer system of claim 13, wherein the predicting, by the machine learning model(s), the execution time for the one or more database queries uses regression analysis based on time series data.

### Claim 16 (depends on 10)

The computer system of claim 10, further comprising identifying a list of data structures for which the update urgency index is to be ascertained, the one or more data structures being in the list of data structures, the list of data structures including any data structures with major data changes since a last updating of quantitative data, any data structures with no quantitative data, and any data structures with stale quantitative data.

### Claim 17

A computer-implemented method of facilitating processing within a computing environment, the computer-implemented method comprising:
training one or more machine learning models to facilitate adaptively updating quantitative data in a database system;
ascertaining an update urgency index for updating quantitative data for one or more data structures of the database system, the update urgency index being representative of an urgency for updating the quantitative data for the data structure(s) and being based, at least in part, on real-time query metrics;
adaptively updating, using the machine learning model(s), the quantitative data, the adaptively updating being based, at least in part, on the ascertained update urgency index; and
optimizing processing of a database query in the database system using, at least in part, the adaptively updated quantitative data.

### Claim 18 (depends on 17)

The computer-implemented method of claim 17, wherein the real-time query metrics comprise a volume and frequency of database queries to the data structure(s) in the database system.

### Claim 19 (depends on 17)

The computer-implemented method of claim 17, wherein the real-time query metrics are based, at least in part, on collecting M top-most database queries to the data structure(s) in order of processor usage for a time interval and estimating therefor a deviation in usage from an estimated processor usage for the M top-most database queries, the estimated deviation being representative of a performance deviation of the M top-most database queries, where M≥1.

### Claim 20 (depends on 10)

The computer-implemented method of claim 10, further comprising predicting, via the machine learning model(s), an execution time of one or more database queries to the data structure(s), wherein the real-time query metrics are based on a difference between the predicted execution time of the one or more database queries to the data structure(s) in comparison to an actual execution time of the one or more database queries to the data structure(s).
