Predicting aggregate value of objects representing potential transactions based on potential transactions expected to be created

Patent metadata

Field	Value
Publication number	US-11651237-B2
Application number	US-201715721346-A
Country	US
Kind code	B2
Filing date	Sep 29, 2017
Priority date	Sep 30, 2016
Publication date	May 16, 2023
Grant date	May 16, 2023

Abstract

An online system stores objects representing potential transactions of an enterprise. The online system uses predictor models to determine an aggregate score based on values of the objects associated with a time interval, for example, a month. Each object is configured to take one of a plurality of states. The online system stores historical data describing activities associated with potential transaction objects and uses the stored data for generating the predictor models. The online system categorizes the objects into bins based on states of the objects. The online system may generate different predictions for each category. The online system may use machine learning based models as predictor models. The online system extracts features describing potential transaction objects and provides these as input to the predictor model.

First claim

We claim:

1. A computer-implemented method comprising:
for a tenant of a multi-tenant online system, storing, a plurality of objects, wherein an object represents a potential transaction and is associated with a value of the potential transaction, the object having one of a plurality of object categories, the object configured to perform category transitions responsive to changes in data associated with the object;
extracting features of a set of objects that were previously processed and interacted with by one or more users, the features including:
a number of changes to a category of an object of the set since the object was created,
a rate of category changes applied to an object of the set, and
an age of an object of the set, wherein the age of the object indicates a time period since the object was created;

training a plurality of linear regression models for a set of object categories, a sequence of sub-intervals in a time interval, and the tenant of the multi-tenant online system based on the extracted features, each of the plurality of linear regression models associated with one of the object categories in the set of categories, one of the sub-intervals in the sequence, and the tenant of the multi-tenant online system, wherein training each of the plurality of linear regression models comprises using a training dataset including extracted features associated with one of the object categories in the set of categories, one of the sub-intervals in the sequence, and the tenant of the multi-tenant online system;
identifying a set of objects from the plurality of objects, the identified set of objects representing potential transactions that are expected to close before the end of the time interval;
selecting, a subset of trained linear regression models from the plurality of trained linear regression models, the subset including trained linear regression models for the tenant, for one or more object categories, and for a position of a current sub-interval in the time interval;
executing the trained linear regression models from the subset to determine total expected values of the identified set of objects for the one or more object categories;
determining a total score for an end of the time interval by aggregating the expected values across the plurality of object categories based on the identified set of objects; and
sending the total score for display by a user interface.

2. The method of claim 1 , wherein an object that was previously processed is stored in a historical data store as a database table, wherein each row of the database table represents a change in the value of the object.

3. The method of claim 1 , wherein each object is represented as a database record.

4. The method of claim 1 , wherein the time interval represents a month and each sub-interval represents a day of the month.

5. A non-transitory computer readable storage medium storing instructions for:
for a tenant of a multi-tenant online system, storing, a plurality of objects, wherein an object represents a potential transaction and is associated with a value of the potential transaction, the object having one of a plurality of object categories, the object configured to perform category transitions responsive to changes in data associated with the object;
extracting features of a set of objects that were previously processed and interacted with by one or more users, the features including:
a number of changes to a category of an object of the set since creation of the object,
a rate of category changes updates applied to an object of the set, and
an age of an object of the set, wherein the age of the object indicates a time period since the object was created;

training a plurality of linear regression models for a set of object categories, a sequence of sub-intervals in a time interval, and the tenant of the multi-tenant online system based on the extracted features, each of the plurality of linear regression models associated with one of the object categories in the set of categories, one of the sub-intervals in the sequence, and the tenant of the multi-tenant online system, wherein training each of the plurality of linear regression models comprises using a training dataset including extracted features associated with one of the object categories in the set of categories, one of the sub-intervals in the sequence, and the tenant of the multi-tenant online system;
identifying a set of objects from the plurality of objects, the identified set of objects representing potential transactions that are expected to close before the end of the time interval;
selecting, a subset of trained linear regression models from the plurality of trained linear regression models, the subset including trained linear regression models for the tenant, for one or more object categories, and for a position of a current sub-interval in the time interval;
executing the trained linear regression models from the subset to determine total expected values of the identified set of objects for the one or more object categories;
determining a total score for an end of the time interval by aggregating the expected values across the plurality of object categories based on the identified set of objects; and
sending the total score for display by a user interface.

6. The non-transitory computer readable storage medium of claim 5 , wherein an object that was previously processed is stored in a historical data store as database table, wherein each row of the database table represents a change in the value of the object.

7. A computer-implemented system of a multi-tenant online system, the system comprising:
a computer processor; and
a computer readable non-transitory storage medium storing instructions thereon, the instructions when executed by a processor cause the processor to perform steps comprising:
for a tenant of a multi-tenant online system, storing, a plurality of objects, wherein an object represents a potential transaction and is associated with a value of the potential transaction, the object having one of a plurality of object categories, the object configured to perform category transitions responsive to changes in data associated with the object;
extracting features of a set of objects that were previously processed and interacted with by one or more users, the features including:
a number of changes to a category of an object of the set since creation of the object,
a rate of category changes applied to an object of the set, and
an age of an object of the set, wherein the age of the object indicates a time period since the object was created;

training a plurality of linear regression models for a set of object categories, a sequence of sub-intervals in a time interval, and the tenant of the multi-tenant online system based on the extracted features, each of the plurality of linear regression models associated with one of the object categories in the set of categories, one of the sub-intervals in the sequence, and the tenant of the multi-tenant online system, wherein training each of the plurality of linear regression models comprises using a training dataset including extracted features associated with one of the object categories in the set of categories, one of the sub-intervals in the sequence, and the tenant of the multi-tenant online system;
identifying a set of objects from the plurality of objects, the identified set of objects representing potential transactions that are expected to close before the end of the time interval;
selecting, a subset of trained linear regression models from the plurality of trained linear regression models, the subset including trained linear regression models for the tenant, for one or more object categories, and for a position of a current sub-interval i

Assignees

• Salesforce Com Inc
• Salesforce Inc

Inventors

• Rickard Jr Scott Thurston
• Balsam Elizabeth Rachel
• Backes Tracy Morgan
• Alexander Zachary

Classifications

• G06N20/00 Primary

  Machine learning · CPC title

• G06N7/01

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

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G06N20/20

  Ensemble learning · CPC title

• G06Q30/0204

  Market segmentation · CPC title