Methods, systems, and computer program product for generating a personalized flow for a software delivery model

Patent metadata

Field	Value
Publication number	US-11270185-B1
Application number	US-201715417758-A
Country	US
Kind code	B1
Filing date	Jan 27, 2017
Priority date	Jan 27, 2017
Publication date	Mar 8, 2022
Grant date	Mar 8, 2022

Abstract

Disclosed are techniques for generating a personalized flow for a software delivery model. These techniques identify user information expressed in natural language for a specific user. One or more user clusters may be determined for the specific user based on at least one user vector representation of a form of the user information. One or more personalized information clusters may be identified for a user cluster of the one or more user clusters based at least in part on the at least one user vector representation. A personalized software application flow may be generated and presented to the specific user by using at least the one or more personalized information clusters for the specific user.

First claim

We claim:

1. A computer implemented method for generating a personalized flow for a software delivery model, comprising:
identifying, at one or more word embedding modules including or coupled with at least one micro-processor of a server coupled to a user computing device through a computer network component, user information of a specific user, the user information expressed in natural language;
determining, at the one or more word embedding modules, one or more user clusters for the specific user, the one or more user clusters each including a plurality of users clustered with the user based in part or in whole upon at least one user vector representation of a form of the user information;
identifying, at the one or more word embedding modules, one or more personalized information clusters for a user cluster of the one or more user clusters based in part or in whole upon the at least one user vector representation; and
generating and presenting a personalized software application flow to the specific user using at least the one or more personalized information clusters for the specific user.

2. The computer implemented method of claim 1 , further comprising:
identifying the user information from the user computing device of the specific user through the computer network component;
normalizing the user information into normalized user information; and
generating, at the one or more word embedding modules, the at least one user vector representation for the normalized user information of the specific user.

3. The computer implemented method of claim 2 , further comprising:
identifying additional user information of a plurality of users;
normalizing the additional user information into additional normalized user information;
generating, at the one or more word embedding modules, a plurality of user vector representations for the additional normalized user information of the plurality of users; and
determining, at the one or more word embedding modules, a plurality of user clusters by processing at least the plurality of user vector representations for the plurality of users.

4. The computer implemented method of claim 3 , further comprising:
identifying characteristic information of a group of users in a user cluster of the one or more user clusters; and
transforming the characteristic information of the group of users in the user cluster into a plurality of characteristic vector representations at least by normalizing the characteristic information into normalized characteristic information and vectorizing the normalized characteristic information.

5. The computer implemented method of claim 4 , further comprising:
determining one or more characteristic clusters for the group of users based in part or in whole upon the plurality of characteristic vector representations; and
identifying, at the one or more word embedding modules, the one or more personalized information clusters for the specific user from the one or more characteristic clusters for the specific user.

6. The computer implemented method of claim 5 , further comprising:
verifying the one or more personalized information clusters by presenting at least one interview session or a chat session to a user interface of the software delivery model.

7. The computer implemented method of claim 6 , further comprising:
identifying additional natural language information of the specific user;
normalizing the additional natural language information into normalized additional natural language information;
transforming the normalized natural language information into one or more additional vector representations; and
modifying the one or more personalized information clusters by determining an addition or removal of at least one personalized information cluster using at least the one or more additional vector representations for the addition natural language information.

8. The computer implemented method of claim 1 , further comprising:
normalizing user information of a plurality of users stored in a central repository into normalized user information; and
reducing a size of the user information or the normalized user information by applying one or more data reduction techniques to the user information or the normalized user information.

9. The computer implemented method of claim 1 , further comprising:
identifying a plurality of analogical reasoning tasks comprising one or more vector arithmetic operations, one or more additive compositionality operations, or one or more combinations of the one or more vector arithmetic operations and one or more additive compositionality operations;
determining one or more training instances with at least the plurality of analogical reasoning tasks; and
training the one or more word embedding modules by executing the one or more training instances for the one or more word embedding modules in a supervised, unsupervised, or reinforcement learning environment.

10. The computer implemented method of claim 1 , further comprising:
identifying one or more corpora including user information and characteristic information of a plurality of users;
identifying or determining a plurality of user vector representations for the user information in the one or more corpora; and
identifying or determining a plurality of user clusters based in part or in whole upon the plurality of user vector representations for the user information in the one or more corpora.

11. The computer implemented method of claim 10 , further comprising:
identifying or determining a plurality of characteristic vector representations for the characteristic information in the one or more corpora; and
identifying or determining a plurality of characteristic clusters for at least one user cluster of the plurality of user clusters based in part or in whole upon the plurality of characteristic vector representations for the characteristic information in the one or more corpora.

12. The computer implemented method of claim 11 , further comprising:
identifying one or more objective tokens;
transforming the one or more objective tokens into one or more objective vector representations at the one or more word embedding modules;
determining similarity scores between the one or more objective vector representations and at least some of the plurality of characteristic vector representations;
identifying one or more target user clusters based in part or in whole upon the similarity scores; and
aggregating at least one target user cluster of the one or more target user clusters and corresponding characteristic information into an aggregated result set.

13. The computer implemented method of claim 1 , generating and presenting the personalized software application flow comprising:
invoking a software flow configuration module in the software delivery model;
identifying a default software application flow comprising a set of default flow nodes for the software application delivery model; and
modifying the default software flow at least by adding one or more flow nodes that correspond to the one or more personalized information clusters to the default software application flow, wherein the one or more flow nodes includes an interview screen or a chat session encompassing information about the one or more personalized information clusters.

14. The computer implemented method of claim 13 , generating and presenting the personalized software application flow further comprising:
modifying the default software application flow at least by removing one or more default flow nodes from the default software application flow.

15. A system

Assignees

• Intuit Inc

Inventors

• Ouyang Peter P
• Rivera Christopher R
• Cessna Joseph B

Classifications

• G06N3/044

  Recurrent networks, e.g. Hopfield networks · CPC title

• G06N3/048

  Activation functions · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/092

  Reinforcement learning · CPC title