# US 2018/0322130 A1 · Data insight discovery using a clustering technique

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

- Number: US 2018/0322130 A1
- Document ID: US20180322130A1
- Status: Published application
- Kind: A1
- Application number: 15/584,400
- Filed: 2017-05-02
- Published: 2018-11-08
- Earliest claimed: 2017-05-02
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Damir Spisic; Jing Xu
- Web page: https://patentgrep.com/patent/US20180322130A1

## CPC

- G06F 16/355: Creation or modification of classes or clusters
- G06F 16/2246: {Trees, e.g. B+trees}
- G06F 16/287: {Visualization; Browsing}
- G06F 16/313: {Selection or weighting of terms for indexing}
- G06F 16/358: Browsing; Visualisation therefor
- G06F 18/2115: by evaluating different subsets according to an optimisation criterion, e.g. class separability, forward selection or backward elimination
- G06F 18/2321: using statistics or function optimisation, e.g. modelling of probability density functions
- G06F 16/00: Information retrieval; Database structures therefor; File system structures therefor
- G06F 16/322: {Trees}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Disclosed aspects relate to data insight discovery using a clustering technique. A set of data may be compressed based on a set of proximity values with respect to a set of predictors to assemble a set of sub-clusters. A set of subgroups may be established by merging a plurality of individual sub-clusters of the set of sub-clusters using a tightness factor. A subset of the subgroups may be selected based on a selection criterion. A set of insight data which indicates a profile of the subset of the set of subgroups with respect to the set of data may be compiled for the subset of the set of subgroups.

## Claims

### Claim 1

A computer-implemented method for data insight discovery using a clustering technique, the method comprising:
compressing, to assemble a set of sub-clusters, a set of data based on a set of proximity values with respect to a set of predictors;
establishing, by merging a plurality of individual sub-clusters of the set of sub-clusters, a set of subgroups using a tightness factor;
selecting, based on a selection criterion, a subset of the set of subgroups; and
compiling, for the subset of the set of subgroups, a set of insight data which indicates a profile of the subset of the set of subgroups with respect to the set of data.

### Claim 2 (depends on 1)

The method of claim 1, further comprising:
constructing, to compress the set of data, a set of Cluster Feature (CF) trees using a sequential clustering technique to scan the set of data on a record-by-record basis.

### Claim 3 (depends on 1)

The method of claim 1, further comprising:
establishing, by merging the plurality of individual sub-clusters of the set of sub-clusters, the set of subgroups using a homogeneity factor with respect to the set of predictors.

### Claim 4 (depends on 3)

The method of claim 3, further comprising:
evaluating, to establish the set of subgroups using the homogeneity factor with respect to the set of predictors, the plurality of individual sub-clusters of the set of sub-clusters based on an attribute comparison with one or more other individual sub-clusters of the set of sub-clusters.

### Claim 5 (depends on 1)

The method of claim 1, further comprising:
comparing a first tightness factor for a first individual sub-cluster with a tightness threshold;
comparing a second tightness factor for a second individual sub-cluster with the tightness threshold;
ascertaining achievement of the tightness threshold by both the first and second tightness factors for the first and second individual sub-clusters, wherein the plurality of individual sub-clusters includes the first and second individual sub-clusters; and
merging the plurality of individual sub-clusters of the set of sub-clusters.

### Claim 6 (depends on 1)

The method of claim 1, further comprising:
configuring the selection criterion to include a subgroup size factor.

### Claim 7 (depends on 1)

The method of claim 1, further comprising:
configuring the selection criterion to include a subgroup tightness factor.

### Claim 8 (depends on 1)

The method of claim 1, further comprising:
configuring the selection criterion to include a subgroup isolation index.

### Claim 9 (depends on 1)

The method of claim 1, further comprising:
finding, to select the subset of the set of subgroups, one or more data clusters without a reliance on decision rules that define subspaces; and
determining, based on a weightiness factor which indicates a relative threshold difference of the subset of the set of subgroups with respect to the set of data, the set of insight data.

### Claim 10 (depends on 1)

The method of claim 1, further comprising:
detecting the set of sub-clusters; and
identifying a set of sub-cluster pairs of the set of sub-clusters.

### Claim 11 (depends on 10)

The method of claim 10, further comprising:
analyzing the set of sub-cluster pairs of the set of sub-clusters; and
choosing, based on a proximity factor in an input space, a chosen pair of the set of sub-cluster pairs of the set of sub-clusters.

### Claim 12 (depends on 11)

The method of claim 11, further comprising:
evaluating, using a target space tightness factor, a merge operation with respect to the chosen pair of the set of sub-cluster pairs of the set of sub-clusters.

### Claim 13 (depends on 11)

The method of claim 11, further comprising:
evaluating, using an input space homogeneity factor, a merge operation with respect to the chosen pair of the set of sub-cluster pairs of the set of sub-clusters.

### Claim 14 (depends on 11)

The method of claim 11, further comprising:
evaluating, using both a target space tightness factor and an input space homogeneity factor, a merge operation with respect to the chosen pair of the set of sub-cluster pairs of the set of sub-clusters; and
merging, based on evaluating the merge operation, the chosen pair of the set of sub-cluster pairs of the set of sub-clusters.

### Claim 15 (depends on 1)

The method of claim 1, further comprising:
configuring the clustering technique to include a hierarchical-oriented clustering technique that uses the tightness factor to indicate an appropriateness of a merge operation.

### Claim 16 (depends on 15)

The method of claim 15, further comprising:
determining the tightness factor based on a Euclidean distance.

### Claim 17 (depends on 1)

The method of claim 1, further comprising:
executing, in a dynamic fashion to streamline data insight discovery using the clustering technique, each of:
the compressing,
the establishing,
the selecting, and
the compiling.

### Claim 18 (depends on 1)

The method of claim 1, further comprising:
executing, in an automated fashion without user intervention, each of:
the compressing,
the establishing,
the selecting, and
the compiling.

### Claim 19

A system for data insight discovery using a clustering technique, the system comprising:
a memory having a set of computer readable computer instructions, and
a processor for executing the set of computer readable instructions, the set of computer readable instructions including:
compressing, to assemble a set of sub-clusters, a set of data based on a set of proximity values with respect to a set of predictors;
establishing, by merging a plurality of individual sub-clusters of the set of sub-clusters, a set of subgroups using a tightness factor;
selecting, based on a selection criterion, a subset of the set of subgroups; and
compiling, for the subset of the set of subgroups, a set of insight data which indicates a profile of the subset of the set of subgroups with respect to the set of data.

### Claim 20

A computer program product for data insight discovery using a clustering technique, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, wherein the computer readable storage medium is not a transitory signal per se, the program instructions executable by a processor to cause the processor to perform a method comprising:
compressing, to assemble a set of sub-clusters, a set of data based on a set of proximity values with respect to a set of predictors;
establishing, by merging a plurality of individual sub-clusters of the set of sub-clusters, a set of subgroups using a tightness factor;
selecting, based on a selection criterion, a subset of the set of subgroups; and
compiling, for the subset of the set of subgroups, a set of insight data which indicates a profile of the subset of the set of subgroups with respect to the set of data.
