Enhancement of machine learning techniques for an electronic message system

• US 9,582,770 B2
• Filed: 04/11/2016
• Issued: 02/28/2017
• Est. Priority Date: 04/29/2015
• Status: Active Grant

• Current Assignee: XANT Inc.
• Original Assignee: Insidesales.com
• Inventors: Zeng, Xinchuan
• Primary Examiner(s): CHEN, ALAN S
• Application Number:
  US15/096,192
• Publication Number:
  US 20160321562A1
• Time in Patent Office: 323 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/107 Computer-aided management of electronic mailing [e-mailing]
• CPC Class Codes: H04L 51/02 using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages
• CPC Class Codes: H04L 51/214 using selective forwarding

Abstract

Techniques are described herein for classifying an electronic message with a particular project from among a plurality of projects. In some embodiments, first and second users associated with the electronic message are identified, and one or more first projects associated with the first user and one and more second projects associated with the second user are determined. Projects that are in common between the first projects and the second projects are determined. When only a single project is in common, the electronic message is associated with the single project. When more than a single project is in common, features associated with each of the projects found to be in common are analyzed by a machine learning model to determine the most likely project to associate with the electronic message from among the projects found to be in common.

Claims

1. A computer system comprising:

a data store that stores a machine learning model and data that indicates which users of a plurality of users are associated with which projects of a set of projects;

one or more computing devices, operatively coupled to the data store, and programmed to;

identify a first user and a second user associated with a first electronic message;

based on the data from the data store, determine a first subset of projects, from the set of projects, that are associated with the first user;

based on the data from the data store, determine a second subset of projects, from the set of projects, that are associated with the second user;

compare the first subset of projects to the second subset of projects to determine a third subset of projects that are common to both the first subset of projects and the second subset of projects;

wherein the third subset of projects includes at least two projects that are in both the first subset of projects and the second subset of projects;

wherein the third subset of projects includes less than all of the set of projects;

collect one or more first features associated with each project of the third subset of projects;

for only those projects that belong to the third subset of projects, generate a project score that reflects likelihood that the first electronic message is related to the project, based at least in part on;

the machine learning model, and

the one or more first features associated with the project;

determine that a particular project has a higher project score than any other project in the third subset of projects; and

in response to determining that the particular project has a higher project score than any other project in the third subset of projects, associate the first electronic message with the particular project.

2. The computer system of claim 1, wherein the one or more features associated with the project comprise one or more of:

goods or services associated with the project;

type of goods or services associated with the project;

persons involved in the project;

size of the project;

price associated with the project;

or

description of the project.

3. The computer system of claim 1, wherein:

the one or more computing devices are further programmed to collect one or more features associated with the first electronic message; and

generating the project score is further based at least in part on the one or more features associated with the first electronic message.

4. The computer system of claim 3, wherein the one or more features associated with the first electronic message comprise one or more of:

content of the first electronic message;

title of the first electronic message;

date the first electronic message was sent or received;

attachments associated with the first electronic message;

domain information of the first electronic message;

or

meta-data of the first electronic message.

5. The computer system of claim 3, wherein:

the one or more computing devices are further programmed to determine a level of similarity between at least one of the one or more features associated with the project and at least one of the one or more features associated with the first electronic message; and

generating the project score is further based at least in part on the determined level of similarity.

6. The computer system of claim 5, wherein the one or more computing devices are further programmed to send the level of similarity to the machine learning model.

7. The computer system of claim 1, wherein generating the project score is further based at least in part on data from the data store indicating that the first or second user was previously identified as belonging to the particular project.

8. The computer system of claim 1, wherein generating the project score is further based at least in part on data from the data store indicating that the first and second users were previously identified as being associated with a second electronic message associated with the particular project.

9. The computer system of claim 1, wherein generating the project score is further based at least in part on data from the data store indicating a number of times the first user was previously associated with an electronic message associated with the particular project.

10. The computer system of claim 1, wherein generating the project score is further based at least in part on data from the data store indicating a number of times the second user was previously associated with an electronic message associated with the particular project.

11. A method comprising:

obtaining data that indicates which users of a plurality of users are associated with which projects of a set of projects;

identifying a first user and a second user associated with a first electronic message;

based on the data from the data store, determining a first subset of projects, from the set of projects, that are associated with the first user;

based on the data from the data store, determining a second subset of projects, from the set of projects, that are associated with the second user;

comparing the first subset of projects to the second subset of projects to determine a third subset of projects that are common to both the first subset of projects and the second subset of projects;

wherein the third subset of projects includes at least two projects that are in both the first subset of projects and the second subset of projects;

wherein the third subset of projects includes less than all of the set of projects;

collecting one or more first features associated with each project of the third subset of projects;

for only those projects that belong to the third subset projects, generating a project score that reflects likelihood that the first electronic message is related to the project, based at least in part on;

a machine learning model, and

the one or more first features associated with the project;

determining that a particular project has a higher project score than any other project in the third subset of projects; and

in response to determining that the particular project has a higher project score than any other project in the third subset of projects, associating the first electronic message with the particular project;

wherein the method is performed by one or more computing devices.

12. The method of claim 11, wherein the one or more features associated with the project comprise one or more of:

goods or services associated with the project;

type of goods or services associated with the project;

persons involved in the project;

size of the project;

price associated with the project;

or

description of the project.

13. The method of claim 11, further comprising:

collecting one or more features associated with the first electronic message; and

wherein the project score is further based at least in part on the one or more features associated with the first electronic message.

14. The method of claim 13, wherein the one or more features associated with the first electronic message comprise one or more of:

content of the first electronic message;

title of the first electronic message;

date the first electronic message was sent or received;

attachments associated with the first electronic message;

domain information of the first electronic message;

or

meta-data of the first electronic message.

15. The method of claim 13, further comprising:

determining a level of similarity between at least one of the one or more features associated with the project and at least one of the one or more features associated with the first electronic message;

wherein generating the project score is further based at least in part on the determined level of similarity.

16. The method of claim 15, further comprising sending the level of similarity to the machine learning model.

17. The method of claim 11, wherein generating the project score is further based at least in part on data from the data store indicating that the first or second user was previously identified as belonging to the particular project.

18. The method of claim 11, wherein generating the project score is further based at least in part on data from the data store indicating that the first and second users were previously identified as being associated with a second electronic message associated with the particular project.

19. The method of claim 11, wherein generating the project score is further based at least in part on data from the data store indicating a number of times the first user was previously associated with an electronic message associated with the particular project.

20. The method of claim 11, wherein generating the project score is further based at least in part on data from the data store indicating a number of times the second user was previously associated with an electronic message associated with the particular project.