Systems and methods for predictive coding

• US 9,595,005 B1
• Filed: 03/20/2013
• Issued: 03/14/2017
• Est. Priority Date: 05/25/2010
• Status: Active Grant

• Current Assignee: Open Text Holdings, Inc. (Open Text Corporation)
• Original Assignee: Recommind Incorporated (Open Text Corporation)
• Inventors: Puzicha, Jan, Vranas, Steve
• Primary Examiner(s): Sitiriche, Luis
• Application Number:
  US13/848,023
• Time in Patent Office: 1,455 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/93 Document management systems
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/10 using kernel methods, e.g. support vector machines [SVM]
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06N 5/048 Fuzzy inferencing
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

Systems and methods for analyzing documents are provided herein. A plurality of documents and user input are received via a computing device. The user input includes hard coding of a subset of the plurality of documents, based on an identified subject or category. Instructions stored in memory are executed by a processor to generate an initial control set, analyze the initial control set to determine at least one seed set parameter, automatically code a first portion of the plurality of documents based on the initial control set and the seed set parameter associated with the identified subject or category, analyze the first portion of the plurality of documents by applying an adaptive identification cycle, and retrieve a second portion of the plurality of documents based on a result of the application of the adaptive identification cycle test on the first portion of the plurality of documents.

Claims

1. A method for analyzing a plurality of documents, comprising:

receiving the plurality of documents via a computing device;

executing instructions stored in memory, wherein execution of the instructions by a processor generates an initial set based on random sampling of a subset of the plurality of documents;

receiving a user input from the computing device, the user input based on an identified subject or category;

reviewing the initial set and the user input to determine at least one seed set parameter associated with the identified subject or category, and

coding a first portion of the plurality of documents, based on the initial set and the at least one seed set parameter associated with the identified subject or category.

2. The method of claim 1, further comprising filtering the plurality of documents to produce the subset of the plurality of documents.

3. The method of claim 1, further comprising coding a second portion of the plurality of documents resulting from an application of user analysis and an adaptive identification cycle.

4. The method of claim 3, further comprising adding the coded second portion of the plurality of documents to the initial set.

5. The method of claim 1, wherein receiving the user input includes a validation of the initial set.

6. The method of claim 1, wherein the random sampling is on a static basis, and wherein generating the initial set is further based on random sampling of the subset of the plurality of documents on a rolling load basis.

7. The method of claim 2, wherein filtering further comprises culling the plurality of documents based on metadata.

8. The method of claim 1, wherein receiving the user input from the computing device comprises a designation corresponding to key documents of the initial set.

9. The method of claim 1, further comprising analyzing the first portion of the plurality of documents by applying an adaptive identification cycle, the adaptive identification cycle being based on at least one of the initial set, user validation of the coding of the first portion of the plurality of documents, and confidence threshold validation.

10. The method of claim 9, further comprising retrieving a second portion of the plurality of documents based on a result of the application of the adaptive identification cycle on the first portion of the plurality of documents.

11. The method of claim 1, further comprising adding further documents to the plurality of documents on a rolling load basis.

12. The method of claim 1, further comprising:

receiving a second user input via the computing device, the second user input corresponding to a confidence level;

calculating a statistic regarding machine-only accuracy rate; and

comparing a statistic regarding machine coding accuracy rate against the second user input based on a defined confidence interval.

13. The method of claim 1, wherein the step of coding a first portion of the plurality of documents further comprises coding based on probabilistic latent semantic analysis and support vector machine analysis of the first portion of the plurality of documents.

14. The method of claim 1, further comprising applying targeted document identification to the plurality of documents.

15. The method of claim 14, wherein applying further comprises identifying a candidate key document based on one or more search terms.

16. The method of claim 14, wherein applying further comprises identifying a candidate key document based on sampling based on one or more phrases.

17. The method of claim 14, wherein applying further comprises identifying a candidate key document using a location filter.

18. The method of claim 14, wherein applying further comprises identifying a candidate key document based on sampling and extrapolation.

19. The method of claim 1, further comprising applying confidence threshold validation testing on one or more of the plurality of documents.

20. The method of claim 19, wherein applying confidence threshold validation testing comprises:

setting a size of a quality control (QC) sample set at a size of the initial set,

creating the QC sample set by random sampling from unreviewed documents, and

reviewing the QC sample set.

21. A method for analyzing a plurality of documents, comprising:

receiving the plurality of documents via a computing device;

executing instructions stored in memory, wherein execution of the instructions by a processor generates an initial set based on random sampling of a subset of the plurality of documents on a static basis and on a rolling load basis;

receiving a user input from the computing device, the user input based on an identified subject or category;

reviewing the initial set and the user input to determine at least one seed set parameter associated with the identified subject or category; and

coding a first portion of the plurality of documents, based on the initial set and the at least one seed set parameter associated with the identified subject or category.

22. The method of claim 21, further comprising applying confidence threshold validation testing on one or more of the plurality of documents.

23. The method of claim 22, wherein applying confidence threshold validation testing comprises:

setting a size of a quality control (QC) sample set at a size of the initial set,

creating the QC sample set by random sampling from unreviewed documents, and

reviewing the QC sample set.

24. The method of claim 21, further comprising adding further documents to the plurality of documents on a rolling load basis.