User behavior segmentation using latent topic detection

Patent metadata

Field	Value
Publication number	US-10242019-B1
Application number	US-201514975654-A
Country	US
Kind code	B1
Filing date	Dec 18, 2015
Priority date	Dec 19, 2014
Publication date	Mar 26, 2019
Grant date	Mar 26, 2019

Abstract

The features relate to artificial intelligence directed compression of user event data based on complex analysis of user event data including latent feature detection and clustering. Further features are described for reducing the size of data transmitted during event processing data flows and devices such as card readers or point of sale systems. Machine learning features for dynamically determining an optimal compression as well as identifying targeted users and providing content to the targeted users based on the compressed data are also included.

First claim

What is claimed is:

1. A method of artificial intelligence guided segmentation of event data, the method comprising:
accessing, from a data store, a plurality of event records associated with respective users of a plurality of users, wherein a first plurality of event records associated with a first user are stored using a first quantity of storage;
accessing an event categories data structure indicating a plurality of event categories and, for each event category, attribute criteria usable to identify events associated with respective event categories;
for the event records,
identifying one or more attributes of the event record,
comparing the identified one or more attributes of the event record to the attribute criteria of respective event categories, and
based on said comparing, assigning, to the event record, an event category having attribute criteria matching the identified one or more attributes of the event record;

generating, for the first user, first compressed event data using the event records associated with the first user and a latent feature identification model, wherein the latent feature identification model takes the event records for the first user and the event categories assigned thereto as an input, and provides association values for the first user for respective event topics identified by the first compressed event data,
wherein first compressed event data associated with the first user is stored using a second quantity of storage, the second quantity of storage being less than the first quantity of storage for storing the event records of the first user;
assigning the first user to one of a plurality of data clusters included in a clustering model using the first compressed event data for the first user; and
generating, for the first user, second compressed event data using a comparison between the first compressed event data for the first user and an average latent feature identification value for a latent feature included in the data cluster to which the first user has been assigned, wherein the second compressed event data associated with the first user is stored using a third quantity of storage, the third quantity of storage being less than the second quantity of storage.

2. The method of claim 1 , wherein assigning the first user to one of the data clusters comprises:
identifying center points for each data cluster included in the clustering model;
generating an association strength for each latent feature included in the first compressed event data for the first user for each data cluster, the association strength indicating a degree of association between the first compressed event data for a user and respective data cluster center points; and
identifying the one of the data clusters as having the highest association strength for the first user from amongst the data clusters included in the clustering model.

3. The method of claim 2 , wherein generating the association strength for the first user comprises comparing a latent feature identification value included in the first compressed event record for a latent feature for the first user to the center point.

4. The method of claim 2 , wherein generating the second compressed event data further comprises:
generating a secondary association strength for each latent feature included in the first compressed event data for a user assigned to the data cluster, the secondary association strength indicating a secondary degree of association between the first compressed event data for the user assigned to the data cluster and the secondary center point of the secondary data cluster to which the user is not assigned, wherein the second compressed event data comprises an identifier for the secondary data cluster and the generated secondary association strengths.

5. The method of claim 2 , further comprising:
accessing content data including a content identifier and an indication of a target data cluster of the data clusters;
identifying a plurality of users assigned to the target data cluster;
selecting a target set of users having second compressed event data including generated association strengths indicating a threshold degree of association to the center point of the target data cluster; and
generating an electronic communication to provide to the target set of user profiles, the electronic communication including content indicated by the content identifier.

6. The method of claim 1 , further comprising:
training the latent feature identification model through probabilistic analysis of a plurality of historical event records to identify a target number of topics; and
training the clustering model using a desired compression level indicating a number of data clusters for the clustering model, wherein training the clustering model includes generating a center point for each data cluster using topically compressed historical event data.

7. The method of claim 1 , wherein the latent feature identification model comprises a latent dirichlet allocation model.

8. A method of compressing transaction data, the method comprising:
receiving a plurality of transaction records each identifying a transaction by one of a plurality of users;
assigning a category to each of the plurality of transaction records;
generating first compressed transaction records using a latent feature identification model, wherein the latent feature identification model takes the transaction records for the one of the plurality of users and categories assigned thereto as an input, and provides association values for the one of the plurality of users for respective topics identified in the first compressed event data;
identifying a clustering compression model for the one of the plurality of users; and
generating second compressed transaction records using the first compressed transaction records and the clustering compression model.

9. The method of claim 8 , wherein generating a first compressed transaction record for the one of the plurality of users comprises receiving association strengths for each topic identified by the latent feature identification model for a set of transactions for the one of the plurality of users.

10. The method of claim 8 , further comprising:
receiving a compression configuration indicating a target number of features to identify for an end user; and
training a latent dirichlet allocation model to identify the target number of features using the received plurality of transaction records, wherein the latent feature identification model comprises the latent dirichlet allocation model.

11. The method of claim 8 , wherein each data cluster included in the clustering compression model is associated with at least one latent feature identifiable by the latent feature identification model, and wherein generating the second compressed transaction records comprises:
assigning each user to one of the data clusters using the first compressed transaction records; and
generating the second compressed transaction records for each user using a comparison between the first compressed transaction data for a user and the center point for the cluster to which the user is assigned.

12. The method of claim 11 , where generating the second compressed transaction records further comprises:
calculating a secondary center point for a secondary data cluster using first compressed transaction data for each user assigned to the secondary data cluster; and
generating a secondary association strength for each latent feature included in the first compressed transaction data for a user assigned to the data cluster, the secondary association strength indicating a secondary degree of association between t

Assignees

• Experian Inf Solutions Inc

Inventors

• Shan Honghao
• Carpenter Mason L
• Bonin Gregor
• Xiong Shanji
• Dichiara Christer

Classifications

• G06N7/01

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

• G06F40/30

  Semantic analysis · CPC title

• G06F40/216 Primary

  using statistical methods · CPC title

• G06N20/00

  Machine learning · CPC title

• G06F16/23

  Updating · CPC title