# US 2021/0004386 A1 · Data visualization platform for event-based behavior clustering

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

- Number: US 2021/0004386 A1
- Document ID: US20210004386A1
- Status: Published application
- Kind: A1
- Application number: 16/927,902
- Filed: 2020-07-13
- Published: 2021-01-07
- Earliest claimed: 2018-06-11
- Assignees: ODAIA Intelligence Inc.
- Applicants: ODAIA Intelligence Inc.
- Inventors: Periklis Andritsos
- Web page: https://patentgrep.com/patent/US20210004386A1

## CPC

- G06F 16/26: Visual data mining; Browsing structured data
- G06F 16/358: Browsing; Visualisation therefor
- G06F 16/2465: {Query processing support for facilitating data mining operations in structured databases}
- G06F 16/287: {Visualization; Browsing}
- G06F 16/288: {Entity relationship models}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A platform for processing event traces to generate clusters of journey maps for an interactive user display. The clustering can be implemented using a genetic process. The clustering can be implemented using a distance measures. Other clustering techniques can be used. The platform can use the clustering results to generate interactive dynamic visualizations with interactive selectable portions. The platform can integrate processing mining and journey maps to generate interactive dynamic visualizations.

## Claims

### Claim 1

A computer system for generating clusters of event traces, the computer system comprising:
a data storage device for storing event traces, each event trace having attributes that indicate activities over time;
a processor configured to process machine executable instructions to generate visual elements for the interactive interface application by:
generating clusters of the event traces based on a genetic process function, each cluster corresponding to a set of event traces and a representative event trace based on representative attributes of the set of event traces, the genetic process function mapping each event trace to a cluster;
generating the visual elements for an interface application, the visual elements indicating the clusters and, for each cluster, the representative event trace based on the representative attributes, the representative event trace summarizing the set of event traces corresponding to the cluster;
controlling rendering of the interface application at a device to display the visual elements and a plurality of selectable portions; and
responsive to a selection of a selectable portion of the plurality of selectable portions, controlling rendering of the interface application at the device to navigate the clusters.

### Claim 2 (depends on 1)

The system of claim 1 wherein the processor is further configured to implement the genetic process function by:
evaluating initial set representative event traces to generate elite set representative event traces;
generating additional initial representative event traces using a transformation process;
evaluating the additional initial representative event traces to generate additional elite set representative event traces;
continuing the generating and the evaluating until stopping criterion is met; and
using resulting elite representative event traces as the representative event traces for the clusters.

### Claim 3 (depends on 1)

The system of claim 1 wherein, for each event trace, the attributes are represented as a two-dimensional tuple of activities over time and contextual data and the genetic process function can evaluate the two-dimensional tuples of the event traces.

### Claim 4 (depends on 1)

The system of claim 1 wherein the processor computes, for each cluster, the representative event trace as a pattern of activities to summarize the patterns contained within the set of event traces of the cluster.

### Claim 5 (depends on 1)

The system of claim 1 wherein the processor uses contextual information to determine the representative event trace.

### Claim 6 (depends on 1)

The system of claim 1 wherein the processor uses one or more neural networks to process event trace attributes.

### Claim 7 (depends on 1)

The system of claim 1 wherein the processor implements the genetic process function by:
pre-processing the event traces;
determining an initial population that corresponds to a number of the clusters;
evaluating initial representative event traces;
evaluating stop criteria;
implementing a genetic operation to generate additional initial representative event traces; and
outputting a map abstraction for use in generating additional visual elements.

### Claim 8 (depends on 7)

The system of claim 7 wherein the processor is configured to evaluate the initial representative event traces based on quality criteria for fitness, a number of representatives, a contextual distance and an average quality.

### Claim 9 (depends on 7)

The system of claim 7 wherein evaluating the stop criteria further comprises at least one selected from the group of:
stopping after a pre-determined number of generations;
stopping after a pre-determined number of generations have been created without improving the average quality; and
stopping once a threshold has been reached for one of the evaluation criteria.

### Claim 10 (depends on 8)

The system of claim 8 wherein the processor is configured to rank the initial representative event traces based on each representative event trace's average quality.

### Claim 11 (depends on 8)

The system of claim 8 wherein the processor is configured to weigh the quality criteria to generate an overall quality and average the overall quality.

### Claim 12 (depends on 1)

The system of claim 1 wherein the processor is configured to evaluate the initial representative event traces based on internal and external evaluation metrics.

### Claim 13 (depends on 12)

The system of claim 12 wherein at least one of the external evaluation metrics is the Jaccard distance.

### Claim 14 (depends on 12)

The system of claim 12 wherein at least one of the internal evaluation metrics is the mean distance between representative event traces and the set of event traces.

### Claim 15

A system for processing event traces to generate a journey map for an interface application comprising:
a data storage device for storing event traces, each event trace having attributes that indicate activities over time, and each event trace relating to current data and historic data;
a processor configured to process machine executable instructions to generate visual elements for an interactive interface application by:
generating a mapping of process mining elements to the event traces, the mapping having an XML-based set of concepts, the mapping linking process de facto models and de jure models to actual and expected customer experiences;
generating a data structure for hierarchical components of a journey map, each component having an element and an attribute, the journey map having at least one element being a root node of the hierarchical components, the actual and expected customer experiences corresponding to the journey map;
generating visual elements for the components of the journey map;
transmitting the visual elements to an interface application for display on a device.

### Claim 16 (depends on 15)

The system of claim 15 wherein the processor uses neural networks to generate tools for process mining.

### Claim 17 (depends on 15)

The system of claim 15 wherein the interface application is configured to explore a plurality of customer journeys simultaneously.

### Claim 18 (depends on 15)

The system of claim 15 wherein the visual elements correspond to representative journeys.

### Claim 19

A system for processing event traces to generate a process tree for an interface application comprising:
a data storage device for storing event traces, the event traces having attributes that indicate activities over time;
a processor configured to process machine executable instructions to generate visual elements for a map abstractor of an interactive interface application by:
generating a process tree for the event traces by grouping the event traces based on similar attributes, the process tree having nodes corresponding to the event traces, the process tree having leaf nodes;
parsing the process tree;
starting from leaf nodes of the parsed process tree, iteratively generating a prompt at an interface application to merge a set of event traces that belong to a subset of the parsed process tree;
upon receiving confirmation to merge the set of event traces, generating a name for the set of event traces and replacing the nodes corresponding to the set of event traces with a new node indicating the name and a set of representative attributes for the set of event traces;
generating the visual elements for an interface application, the visual elements indicating the process tree and the new nodes, the visual elements representing an abstracted process tree; and
controlling rendering of the interface application at a device to display the visual elements.

### Claim 20 (depends on 19)

The system of claim 19, wherein a process tree is an abstract hierarchical representation of a process model, where the leaves are annotated with activities and all the other nodes are annotated with operators.
