# US 2017/0288989 A1 · Systems and Techniques for Determining Associations Between Multiple Types of Data in Large Data Sets

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

- Number: US 2017/0288989 A1
- Document ID: US20170288989A1
- Status: Published application
- Kind: A1
- Application number: 15/085,210
- Filed: 2016-03-30
- Published: 2017-10-05
- Earliest claimed: 2016-03-30
- Assignees: Adobe Systems Incorporated
- Applicants: Adobe Systems Incorporated
- Inventors: Ritwik Sinha; Said Kobeissi; Michael Young
- Web page: https://patentgrep.com/patent/US20170288989A1

## CPC

- G06T 11/26: Drawing of charts or graphs
- G06F 16/35: Clustering; Classification
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Systems and methods disclosed herein identify multivariate relationships that exist across all types data collected from numerous observed users over one or more networks. Electronic data collected from observed users include categorical data and non-categorical/numeric data. To compare and analyze the collected data, a marketing entity converts the numeric data to categorical data via a binning algorithm, which reduces the numeric data into two or more discrete categories. The marketing entity analyzes the data variables to compute pairwise associations on the collected categorical and numeric data (which has been converted to categorical data). The marketing entity also determines hierarchical clusters to group the pairwise associations of data variables based on the strength of the associations. The pairwise relationships and hierarchical clusters are displayed on a user interface.

## Claims

### Claim 1

A method for enhancing graphical visualizations in computer applications depicting multivariate relationships of electronic data on Internet analytics, the electronic data collected from interactions on one or more websites by a plurality of observed users, the method comprising:
receiving, from a network-based server, electronic data collected from a plurality of observed users, the electronic data distributed across storage units in a distributed computing environment, wherein the electronic data includes a plurality of sets of continuous data and a plurality of sets of categorical data, wherein the distributed computing environment includes a plurality of distributed processing units, each distributed processing unit for processing operations on a different one of the sets of electronic data;
converting the plurality of sets of continuous data to converted categorical data by applying a binning algorithm on the plurality of sets of continuous data; and
configuring a display device to display a graphical indication of pairwise relationships among data variables included in the converted categorical data and the categorical data.

### Claim 2 (depends on 1)

The method of claim 1, wherein converting the plurality of sets of continuous data to converted categorical data comprises:
identifying, by each distributed processing unit, individual maximum and minimum values from each of the sets of continuous data and determining an overall minimum value and an overall maximum value from the individual maximum and minimum values;
determining two or more bins from the overall maximum value and the overall minimum value, the two or more bins respectively corresponding to two or more categories of data; and
determining how many data points are included in the two or more bins and categorizing individual continuous data points included in the plurality of sets of continuous data by distributing the individual continuous data points into the two or more bins, each bin corresponding to a category in the converted categorical data.

### Claim 3 (depends on 1)

The method of claim 1, wherein configuring the display device to display the graphical indication of pairwise relationships among data variables comprises:
determining pairwise relationships between data variables included in the plurality of sets of categorical data and the converted categorical data by determining uncertainty coefficients between each of the data variables;
determining, from the pairwise relationships, groupings of the data variables included in the plurality of sets of categorical data and the converted categorical data by strength of associations between the data variables, and
generating the graphical indication, wherein the graphical indication indicates the strength of relationships between at least some of the data variables.

### Claim 4 (depends on 1)

The method of claim 1, wherein the graphical indication of pairwise relationships between the data variables includes a heatmap of a computational association matrix, wherein each cell in the computational association matrix indicates the strength of the correlation between two of the data variables, and wherein the cells in the computational association matrix are organized in one or more clusters of shaded regions, the one or more clusters indicating groups of the data variables having strong relationships with each other.

### Claim 5 (depends on 1)

The method of claim 1, wherein the graphical indication of relationships between all of the data variables includes a chord diagram depicting the strength of the associations between the data variables via thickness of connecting arcs.

### Claim 6 (depends on 1)

The method of claim 1, wherein the electronic data collected from the plurality of observed users includes aggregated session-level data and individual-level data, wherein the session-level data includes a group of interactions with one or more of the data variables during single login sessions from the plurality of observed users and wherein the individual-level data includes a different group of interactions with one or more of the data variables during multiple login sessions from the plurality of observed users.

### Claim 7 (depends on 6)

The method of claim 6, further comprising:
receiving a selection from a user interface specifying a combination of the session-level data and the individual-level data to use for aggregation.

### Claim 8 (depends on 1)

The method of claim 1, further comprising:
reducing an amount of cardinality of the plurality of sets of categorical variables and of the converted categorical variables by grouping infrequently occurring data points into a separate category.

### Claim 9 (depends on 1)

The method of claim 1, wherein the electronic data comprises Internet analytics data describing demographic information, location information, and Internet usage data collected from interactions on one or more websites by the plurality of observed users.

### Claim 10

A system for enhancing graphical visualizations in computer applications depicting multivariate relationships of electronic data on Internet analytics, the electronic data collected from interactions on one or more websites by a plurality of observed users, the method comprising:
a processor; and
a memory communicatively coupled to the processor, wherein the processor is configured to execute instructions included in the memory device to perform operations comprising:
receiving, from a network-based server, electronic data collected from a plurality of observed users, the electronic data distributed across storage units in a distributed computing environment, wherein the electronic data includes a plurality of sets of continuous data and a plurality of sets of categorical data, and wherein the distributed computing environment includes a plurality of distributed processing units, each distributed processing unit for processing operations on a different one of the sets of electronic data;
converting the plurality of sets of continuous data to converted categorical data by applying a binning algorithm on the plurality of sets of continuous data; and
configuring a display device to display a graphical indication of pairwise relationships among data variables included in the converted categorical data and the categorical data.

### Claim 11 (depends on 10)

The system of claim 10, wherein converting the plurality of sets of continuous data to converted categorical data comprises:
identifying, by each distributed processing unit, individual maximum and minimum values from each of the sets of continuous data and determining an overall minimum value and an overall maximum value from the individual maximum and minimum values;
determining two or more bins from the overall maximum value and the overall minimum value, the two or more bins respectively corresponding to two or more categories of data; and
determining how many data points are included in the two or more bins and categorizing individual continuous data points included in the plurality of sets of continuous data by distributing the individual continuous data points into the two or more bins, each bin corresponding to a category in the converted categorical data.

### Claim 12 (depends on 10)

The system of claim 10, wherein configuring the display device to display the graphical indication of pairwise relationships among data variables comprises:
determining pairwise relationships between data variables included in the plurality of sets of categorical data and the converted categorical data;
determining, from the pairwise relationships, groupings of the data variables included in the plurality of sets of categorical data and the converted categorical data by strength of associations between the data variables; and
generating the graphical indication, wherein the graphical indication indicates the strength of relationships between at least some of the data variables.

### Claim 13 (depends on 10)

The system of claim 10, wherein the graphical indication of pairwise relationships between the data variables includes a heatmap of a computational association matrix, wherein each cell in the computational association matrix indicates the strength of the correlation between two of the data variables, and wherein the cells in the computational association matrix are organized in one or more clusters of shaded regions, the one or more clusters indicating groups of the data variables having strong relationships with each other.

### Claim 14 (depends on 10)

The system of claim 10, wherein the graphical indication of relationships between all of the data variables includes a chord diagram depicting the strength of the associations between the data variables via thickness of connecting arcs.

### Claim 15 (depends on 10)

The system of claim 10, wherein the electronic data collected from the plurality of observed users includes aggregated session-level data and individual-level data, wherein the session-level data includes a group of interactions with one or more of the data variables during single login sessions from the plurality of observed users and wherein the individual-level data includes a different group of interactions with one or more of the data variables during multiple login sessions from the plurality of observed users.

### Claim 16 (depends on 15)

The system of claim 15, wherein the processor is configured to execute instructions included in the memory device to perform operations further comprising:
receiving a selection from a user interface specifying a combination of the session-level data and the individual-level data to use for aggregation.

### Claim 17 (depends on 10)

The system of claim 10, wherein the processor is configured to execute instructions included in the memory device to perform operations further comprising:
reducing an amount of cardinality of the plurality of sets of categorical variables and of the converted categorical variables by grouping infrequently occurring data points into a separate category.

### Claim 18 (depends on 10)

The system of claim 10, wherein the electronic data comprises Internet analytics data describing demographic information, location information, and Internet usage data collected from interactions on one or more websites by the plurality of observed users

### Claim 19

A non-transitory computer-readable medium with program code stored thereon, wherein the program code is executable to perform operations comprising:
receiving, from a network-based server, the electronic data collected from the plurality of observed users, the electronic data including categorical data and continuous data and the electronic data stored across storage units in a distributed computing environment, and wherein the electronic data comprises Internet analytics data describing demographic information, location information, and Internet usage data collected from interactions on one or more websites by the plurality of observed users;
converting the continuous data to converted categorical data, wherein converting the continuous data to converted categorical data comprises:
collecting individual maximum and individual minimum values of individual continuous data points from each data processing unit within the distributed computing environment,
determining an overall minimum value and an overall maximum value from the individual maximum and minimum values,
determining two or more bins from the overall maximum value and the overall minimum value, the two or more bins respectively corresponding to two or more categories of data, and
determining how many data points are included in the two or more bins to categorize the individual data points included in the continuous data into the two or more categories;
determining pairwise relationships between data variables included in the categorical data and the converted categorical data;
determining, from the pairwise relationships, groupings of the data variables included in the categorical data and the converted categorical data by strength of associations between the data variables; and
displaying a graphical indication of relationships between at least some of the data variables.

### Claim 20 (depends on 15)

The non-transitory computer-readable medium of claim 15, wherein the graphical indication of relationships between the data variables includes a heatmap of a computational association matrix, wherein each cell in the computational association matrix indicates the strength of the correlation between two of the data variables, and wherein the cells in the computational association matrix are organized in one or more clusters of shaded regions, the one or more clusters indicating groups of the data variables having strong relationships with each other.
