Method of automated discovery of new topics

• US 9,177,262 B2
• Filed: 12/02/2014
• Issued: 11/03/2015
• Est. Priority Date: 12/02/2013
• Status: Active Grant

• Current Assignee: Finch Computing LLC (Qbase, LLC)
• Original Assignee: Qbase, LLC
• Inventors: Lightner, Scott, Weckesser, Franz, Boddhu, Sanjay, Flagg, Robert
• Primary Examiner(s): Chaki, Kakali
• Assistant Examiner(s): SITIRICHE, LUIS A
• Application Number:
  US14/558,076
• Publication Number:
  US 20150154148A1
• Time in Patent Office: 336 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/93 Document management systems
• CPC Class Codes: G06F 40/00 Handling natural language data speech analysis or synthesis, speech recognition G10L
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G06N 5/04 Inference or reasoning models

Abstract

The present disclosure relates to a method for performing automated discovery of new topics from unlimited documents related to any subject domain, employing a multi-component extension of Latent Dirichlet Allocation (MC-LDA) topic models, to discover related topics in a corpus. The resulting data may contain millions of term vectors from any subject domain identifying the most distinguished co-occurring topics that users may be interested in, for periodically building new topic ID models using new content, which may be employed to compare one by one with existing model to measure the significance of changes, using term vectors differences with no correlation with a Periodic New Model, for periodic updates of automated discovery of new topics, which may be used to build a new topic ID model in-memory database to allow query-time linking on massive data-set for automated discovery of new topics.

Claims

1. A method comprising:

automatically extracting, by a database source computer, from a document corpus, data associated with a plurality of co-occurring topics;

in response to automatically extracting the plurality of co-occurring topics, extracting, by a synchronizing framework computer, a plurality of topic identifiers from the plurality of co-occurring topics;

creating, by the synchronizing framework computer, a master topic computer model for the document corpus from a first plurality of term vectors;

creating, by the synchronizing framework computer, a periodic new topic computer model by comparing topic significance among the plurality of topic identifiers, the periodic new topic computer model including a second plurality of term vectors; and

selecting, by the synchronizing framework computer, one or more new topics by identifying one or more term vectors from the second plurality of term vectors in the periodic new topic computer model that have no correlation with the first plurality of term vectors in the master topic computer model.

2. The method of claim 1 further comprising adding, via the synchronizing framework computer module, the one or more new topics to the master topic computer model.

3. The method of claim 1 further comprising:

receiving, via the database source computer module, an indication of a topic of interest to a user; and

automatically extracting the data associated with the plurality of co-occurring topics based on the topic of interest.

4. The method of claim 1 wherein the master topic computer model is a multi-component extension of a Latent Dirichlet Allocation (MC-LDA) topic model.

5. The method of claim 1 wherein the periodic new topic computer model is a multi-component extension of a Latent Dirichlet Allocation (MC-LDA) topic model.

6. The method of claim 1 wherein comparing topic significance among the plurality of topic identifiers is based on a predetermined significance threshold.

7. A system comprising:

a database source computer module configured to extract data associated with a plurality of co-occurring topics in a document corpus; and

a synchronizing framework computer module configured to;

extract a plurality of topic identifies from the plurality of co-occurring topics;

create a master topic computer model for the document corpus from a first plurality of term vectors;

create a periodic new topic computer model by comparing topic significance among the plurality of topic identifiers, the periodic new topic computer model including a second plurality of term vectors; and

select one or more new topics by identifying one or more term vectors from the second plurality of term vectors in the periodic new topic computer model that have no correlation with the first plurality of term vectors in the master topic computer model.

8. The system of claim 7 wherein the synchronizing framework computer module is further configured to add the one or more new topics to the master topic computer model.

9. The system of claim 7 wherein the database source computer module is further configured to receive an indication of a topic of interest to a user for extraction of the data associated with the plurality of co-occurring topics based on the topic of interest.

10. The system of claim 7 wherein the master topic computer model is a multi-component extension of a Latent Dirichlet Allocation (MC-LDA) topic model.

11. The system of claim 7 wherein the periodic new topic computer model is a multi-component extension of a Latent Dirichlet Allocation (MC-LDA) topic model.

12. The system of claim 7 wherein the synchronizing framework computer module is further configured to compare topic significance among the plurality of topic identifiers based on a predetermined significance threshold.

13. A non-transitory computer readable medium having stored thereon computer executable instructions executed by a processor comprising:

automatically extracting, by a processor executing a database source computer module, from a document corpus data associated with a plurality of co-occurring topics;

in response to automatically extracting the plurality of co-occurring topics, extracting, by the processor executing a synchronizing framework computer module, a plurality of topic identifiers from the plurality of co-occurring topics;

creating, by the processor executing the synchronizing framework computer, a master topic computer model for the document corpus from a first plurality of term vectors;

creating, by the processor executing the synchronizing framework computer, a periodic new topic computer model by comparing topic significance among the plurality of topic identifiers, the periodic new topic computer model including a second plurality of term vectors; and

selecting, by the processor executing the synchronizing framework computer, one or more new topics by identifying one or more term vectors from the second plurality of term vectors in the periodic new topic computer model that have no correlation with the first plurality of term vectors in the master topic computer model.

14. The computer readable medium of claim 13, wherein the instructions further comprise adding, by the processor executing the synchronizing framework computer module, the one or more new topics to the master topic computer model.

15. The computer readable medium of claim 13, wherein the instructions further comprise:

receiving, by the processor executing database source computer module, an indication of a topic of interest to a user; and

automatically extracting the data associated with the plurality of co-occurring topics based on the topic of interest.

16. The computer readable medium of claim 13, wherein the master topic computer model is a multi-component extension of a Latent Dirichlet Allocation (MC-LDA) topic model.

17. The computer readable medium of claim 13, wherein the periodic new topic computer model is a multi-component extension of a Latent Dirichlet Allocation (MC-LDA) topic model.

18. The computer readable medium of claim 13, wherein comparing topic significance among the plurality of topic identifiers is based on a predetermined significance threshold.