Multi-service business platform system having entity resolution systems and methods

Patent metadata

Field	Value
Publication number	US-11775494-B2
Application number	US-202117318737-A
Country	US
Kind code	B2
Filing date	May 12, 2021
Priority date	May 12, 2020
Publication date	Oct 3, 2023
Grant date	Oct 3, 2023

Abstract

The disclosure is directed to various ways of improving the functioning of computer systems, information networks, data stores, search engine systems and methods, and other advantages. Among other things, provided herein are methods, systems, components, processes, modules, blocks, circuits, sub-systems, articles, and other elements (collectively referred to in some cases as the “platform” or the “system”) that collectively enable, in one or more datastores (e.g., where each datastore may include one or more databases) and systems, the creation, development, maintenance, and use of a set of custom objects for use in a wide range of activities, including sales activities, marketing activities, service activities, content development activities, and others, as well as improved methods and systems for sales, marketing and services that make use of such entity resolution systems and methods as well as custom objects.

First claim

What is claimed is:

1. A method, comprising:
receiving object properties of objects to use for filtering objects maintained by a multi-service platform for a client, wherein the objects represent entities;
utilizing, by a machine learning model, the object properties to filter the objects to identify a set of objects related to the object properties, wherein the machine learning module processes feature vectors of the entities to produce reduced entity-specific vectors for entities represented by the set of objects;
generating a prediction providing insight into the set of objects by utilizing the set of objects, wherein the reduced entity specific vectors are organized as a first entity feature matrix that is processed with a second entity feature matrix to generate the prediction as a likelihood that pairs of the entities represented by the set of objects are duplicates; and
performing a deduplication action based upon the prediction.

2. The method of claim 1 , wherein the objects comprise custom objects, and wherein the machine learning model is trained to provide custom object filtering.

3. The method of claim 1 , comprising:
utilizing instances of custom objects to train the machine learning model to process core objects and custom objects.

4. The method of claim 1 , wherein a first object representing a first entity is deleted based upon the first entity being identified as a duplicate of a second entity represented by a second object.

5. The method of claim 1 , comprising:
merging a first object and a second object based upon the first object and the second object representing duplicate entities.

6. The method of claim 1 , comprising:
performing, by the machine learning model, custom object filtering that includes list segmentation.

7. The method of claim 1 , comprising:
searching across custom object types to automate changes in custom objects.

8. The method of claim 1 , comprising:
performing, by the machine learning model, custom object filtering that includes searching across custom object types to automate changes in core objects maintained as preset objects by the multi-service platform.

9. The method of claim 1 , wherein the machine learning model is an entity deduplication model trained to predict duplicate entities.

10. The method of claim 9 , comprising:
training the entity deduplication model to produce an entity-specific vector that minimizes a training error generated by comparing a preconfigured p-merge value for a pair of training entities to a deduplicate likelihood value for the pair of training entities.

11. The method of claim 10 , comprising:
processing a pair of entity-specific vectors for the pair of training entities to generate the deduplicate likelihood value for the pair of training entities.

12. The method of claim 10 , comprising:
deriving the preconfigured p-merge value by performing string matching upon features of the pair of training entities.

13. The method of claim 10 , comprising:
deriving the preconfigured p-merge value by applying heuristics to features of the pair of training entities.

14. A system, comprising:
memory storing instructions; and
a processor that executes the instructions to perform operations comprising:
receiving object properties of objects to use for filtering objects maintained by a multi-service platform for a client, wherein the objects represent entities;
utilizing, by a machine learning model, the object properties to filter the objects to identify a set of objects related to the object properties, wherein the machine learning module processes feature vectors of the entities to produce reduced entity-specific vectors for entities represented by the set of objects;
generating a prediction providing insight into the set of objects by utilizing the set of objects, wherein the reduced entity specific vectors are organized as a first entity feature matrix that is processed with a second entity feature matrix to generate the prediction as a likelihood that pairs of the entities represented by the set of objects are duplicates; and
performing a deduplication action based upon the prediction to the client.

15. The system of claim 14 , wherein a first object representing a first entity is deleted based upon the first entity being identified as a duplicate of a second entity represented by a second object.

16. The system of claim 14 , wherein the operations comprise:
merging a first object and a second object based upon the first object and the second object representing duplicate entities.

17. The system of claim 14 , wherein the operations comprise:
performing, by the machine learning model, custom object filtering that includes list segmentation.

18. The system of claim 14 , wherein the operations comprise:
performing, by the machine learning model, custom object filtering that includes searching across custom object types to automate changes in custom objects.

19. The system of claim 14 , wherein the operations comprise:
performing, by the machine learning model, custom object filtering that includes searching across custom object types to automate changes in core objects.

20. The system of claim 14 , wherein the machine learning model is an entity deduplication model trained to predict duplicate entities.

21. A non-transitory computer readable storage medium storing instructions which when executed by a processor cause the processor to perform operations comprising:
receiving object properties of objects to use for filtering objects maintained by a multi-service platform for a client, wherein the objects represent entities;
utilizing, by a machine learning model, the object properties to filter the objects to identify a set of objects related to the object properties, wherein the machine learning module processes feature vectors of the entities to produce reduced entity-specific vectors for entities represented by the set of objects;
generating a prediction providing insight into the set of objects by utilizing the set of objects, wherein the reduced entity specific vectors are organized as a first entity feature matrix that is processed with a second entity feature matrix to generate the prediction as a likelihood that pairs of the entities represented by the set of objects are duplicates; and
performing a deduplication action based upon the prediction.

Assignees

• Hubspot Inc

Inventors

• Urdiales Hector
• Lagi Marco
• Purcell Stephen J
• Layton Stuart P
• Ash Bryan
• Williams Jared
• Higgs Sophie
• Mceneaney Robert
• Sellberg Dylan
• Perko Anna

Classifications

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0455

  Auto-encoder networks; Encoder-decoder networks · CPC title

• G06F16/2237 Primary

  Vectors, bitmaps or matrices · CPC title

• G06F16/215 Primary

  Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors · CPC title