Method and system for determining behavioral patterns of user groups based on interactions with other user groups using machine learning

Patent metadata

Field	Value
Publication number	US-12499376-B2
Application number	US-202117180418-A
Country	US
Kind code	B2
Filing date	Feb 19, 2021
Priority date	Feb 19, 2021
Publication date	Dec 16, 2025
Grant date	Dec 16, 2025

Abstract

The disclosure describes a method of generating a target profile including the target's sequence of events (SOE) for a task. Such target profile sequence of events is derived from several source group's transactions, where any source group's transactions cannot be shared with other source groups but the derived target group's profile is the only information that is shared. Source-side information is periodically extracted for a plurality of sources that each interact with a plurality of targets. The information includes source stages, resources, and stage transition events for a task with a target. Source information is used to generate a set of normalized stages, and a set of normalized events for transitioning between the stages of the set of normalized stages. An artificial intelligence (AI) model is trained using the source information. The AI model can generate a target profile with target process information inferred using the trained model. The target process information can include the target's identifiers for each stage, an estimated duration of the stage, deliverables for the stage, and one or more stage transition events for the stage.

First claim

What is claimed is:

1 . A computer-implemented method of profiling a user group, the method comprises:
periodically via a first thread executed by a first processor,
obtaining event data from a task database system, the event data including a plurality of events of a plurality of tasks performed between a plurality of source user groups and a plurality of target user groups, wherein the plurality of source user groups have opted-in to sharing their task data and each source user group in the source user groups cannot access information in the task database that was received from another source user group,
wherein each task includes a plurality of stages, wherein each event corresponds to an activity performed by a particular source user group on a particular task with respect to a particular target user group within a predetermined time period in the past,
normalizing the plurality of events and the plurality of stages of the plurality of tasks, comprising:
for each of the source user groups, determining the greatest number of stages in a task process associated with the source user group;
determining an average number of stages across the plurality of source user groups using the determined greatest number of stages for each source user group; then
generating a set of normalized stages having the determined average number of stages;
for each stage in the task process of each source user group, determining one or more events that are required to advance the task process to the next stage in the task process; then
generating the set of normalized events for transitioning to a next stage of the set of normalized stages, using the determined events for each stage in the task process of each source user group; and
for each of one or more stages of the task process of each source user group, matching the stage to a stage in the set of normalized stages by matching an event in the one or more events that are required to advance the task process of the source user group to an event in the set of normalized events that is associated with the matching stage in the set of normalized stages,

extracting a set of task features of the plurality of tasks from the event data, wherein the set of task features identify stages of each task, a transition event of the stage, timing information from one stage to another stage of each task, and a completion rate of the tasks, and
training an artificial intelligence (AI) model using the extracted set of task features as a training set of data, training the AI model comprises mapping the extracted stages of each task and the extracted transition event of the stage to a set of normalized stages and a set of normalized events for transitioning to a next stage of the set of normalized stages; and
via a second thread executed by a second processor,
receiving a request for profiling a first target user group from a client device over a network and the first target user group does not share its data,
in response to the request, retrieving from the task database system a set of real-time event data corresponding to a set of recent events performed by one or more first source user groups on one or more tasks with respect to the first target user group,
applying the AI model to a set of real-time task features extracted from the set of recent events to generate a first user group profile describing a behavioral pattern of the first user target group, the first user group profile comprising:
a projected duration of each stage of a potential task involved by the first target user group, and
a contact information of one or more first target user group personnel associated with the stage of the potential task,

wherein the generated first user group profile is shareable between the plurality of source user groups, and
transmitting the first user group profile to the client device over the network.

2 . The method of claim 1 , wherein the set of task features extracted from the event data comprises a user identifier (ID) of each user of each source user group involved in each stage of each task and a user role of each user.

3 . The method of claim 2 , wherein the set of task features extracted from the event data further comprises a completion rate of past tasks involved by each user of each source user group.

4 . The method of claim 1 , wherein the set of task features extracted from the event data comprises duration of each stage of each of the plurality of tasks.

5 . The method of claim 1 , wherein the set of task features extracted from the event data comprises a size of each task.

6 . The method of claim 1 , wherein the set of task features extracted from the event data comprises a date and time stamp indicating when a piece of information in the event data was updated by a member of a corresponding source user group.

7 . The method of claim 1 , wherein the first target user group profile comprises a probability of advancing to a next stage of the potential task.

8 . The method of claim 1 , wherein the first target user group profile comprises:
a trend in change of duration of a transition time between one or more of the normalized stages; one or more outlier instances in a change of the duration of a transition time between one or more of the normalized stages; or a seasonality of the duration of a transition time between one or more of the normalized stages.

9 . The method of claim 1 , wherein the task database system is configured to store task activity data for the source user group and is accessible by the source user groups, and wherein the first target user group profile is determined without having to contacting the first target user group.

10 . The method of claim 1 , wherein the AI model includes at least one of a hidden Markov model, an echo state network, or a Bayesian network.

11 . A data processing system, comprising:
a processor; and a memory coupled to the processor to store instructions therein for profiling a user group, which when executed by the processor, cause the processor to perform operations, the operations including: periodically via a first thread executed by a first processor,
obtaining event data from a task database system, the event data including a plurality of events of a plurality of tasks performed between a plurality of source user groups and a plurality of target user groups, wherein the plurality of source user groups have opted-in to sharing their task data and each source user group in the source user groups cannot access information in the task database that was received from another source user group,
wherein each task includes a plurality of stages, wherein each event corresponds to an activity performed by a particular source user group on a particular task with respect to a particular target user group within a predetermined time period in the past,
normalizing the plurality of events and the plurality of stages of the plurality of tasks using comprising:
for each of the source user groups, determining the greatest number of stages in a task process associated with the source user group;
determining an average number of stages across the plurality of source user groups using the determined greatest number of stages for each source user group; then
generating a set of normalized stages having the determined average number of stages;
for each stage in the task process of each source user group, determining one or more events that are required to advance the task process to the next stage in the task process; then
generating the set of normalized events for transitioning to a next stage of the set of normalized stages, using the determined events for each stage in the task process

Assignees

• Clari Inc

Inventors

• Rangan Venkat

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06F16/285

  Clustering or classification · CPC title

• G06N20/00

  Machine learning · CPC title

• G06N5/04 Primary

  Inference or reasoning models · CPC title

• G06F16/95 Primary

  Retrieval from the web · CPC title