IP Library Granted Patent US 10,572,576
Granted Patent B1
US 10,572,576 · App. 15/633,239 · Granted Feb 25, 2020

Systems and methods for facilitating data object extraction from unstructured documents

Inventors: Brandon Marc-Aurele (Arlington, VA); John Doyle (Washington, DC)
Assignee: Palantir Technologies Inc.
G06F17/218G06F17/241G06F17/278G06F17/2765
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 10,572,576
App. No.
15/633,239
Granted
Feb 25, 2020
Kind
B1
Abstract

Systems and methods are provided for facilitating data object extraction from unstructured documents. Unstructured documents may include data in an unorganized format, such as raw text. The system may use natural language processing to determine characteristics of the terms used in the unstructured document. The system may prompt a user to select terms from the document corresponding in characteristics to properties of a data object being generated. The user may select terms from the document and the system may generate a data object according to the selected terms.

Claims (61)

1. A system for extracting object data from an unstructured document:

one or more processors and a memory storing instructions that, when executed by the one or more processors, cause the system to:

receive an unstructured document comprising a plurality of terms, each term including at least one word, wherein the unstructured document is not organized according to an ontology;

assign a term classification to the plurality of terms of the unstructured document via natural language processing;

create a tagging template for the unstructured document by generating a plurality of tagging elements of the tagging template according to a required data object structure;

obtain the tagging template for the unstructured document, the tagging template including the plurality of tagging elements, each tagging element having an element classification, the tagging template being obtained based on a particular ontology and a description of the unstructured document, the plurality of tagging elements of the tagging template corresponding to one or more properties of the particular ontology;

receive, from a user via an interface, a plurality of selected terms corresponding to the plurality of tagging elements, wherein the term classifications of the selected terms matches the element classification of the corresponding tagging elements;

generate, from the unstructured document based on the plurality of selected terms corresponding to the plurality of tagging elements, a data object organized according to the particular ontology, the generated data object conforming to the required data object structure;

determine a selected tagging element from the plurality of tagging elements;

prompt to the user, via the interface, suggested terms from the plurality of selected terms having a term classification matching the element classification of the selected tagging element; and

receive, from the user, a narrative structure indicative of at least one relationship between the plurality of tagging elements of the tagging template.

2. The system of claim 1 , wherein the system is further caused to provide, to the user via the interface, a visual display of the plurality of tagging elements and the narrative structure indicating the relationship between the plurality of tagging elements.

3. The system of claim 1 , wherein the system is further caused to:

determine a selected tagging element from the plurality of tagging elements; and

provide, to the user, a prompt providing information about at least one characteristic of the selected tagging element.

4. The system of claim 1 , wherein the system is further caused to:

determine a first selected tagging element from the plurality of tagging elements;

identify the first selected tagging element to the user;

receive a first selected term corresponding to the first selected tagging element according to user input;

determine a second selected tagging element from the plurality of tagging elements;

identify the second selected tagging element to the user;

receive a second selected term corresponding to the second selected tagging element according to user input,

wherein the plurality of selected terms includes at least the first selected term and the second selected term.

5. The system of claim 1 , wherein the system is further caused to:

receive, from a user, a selection of a second data object; and

store a data link between the generated data object and the selected second data object.

6. The system of claim 1 , wherein the system is further caused to:

receive, from a second user, a second plurality of selected terms corresponding to the plurality of tagging elements, wherein the term classifications of the selected terms matches the element classification of the corresponding tagging elements;

receive, from a third user, a third plurality of selected terms corresponding to the plurality of tagging elements, wherein the term classifications of the selected terms matches the element classification of the corresponding tagging elements; and

wherein to generate the data object from the unstructured document the system is caused to generate the data object according to the plurality of selected terms, the second plurality of selected terms, and the third plurality of selected terms.

7. The system of claim 1 , wherein to assign the term classification to the plurality of terms, the system is further caused to assign to each of the plurality of terms at least one of a part of speech and a grammatical role.

8. A method for extracting object data from an unstructured document, the method being performed on a computer system having one or more physical processors programmed with computer program instructions that, when executed by the one or more physical processors, cause the computer system to perform the method, the method comprising:

receiving, by the computer system, an unstructured document comprising a plurality of terms, each term including at least one word, wherein the unstructured document is not organized according to an ontology;

assigning, by the computer system, a term classification to the plurality of terms of the unstructured document via natural language processing;

creating a tagging template for the unstructured document by generating a plurality of tagging elements of the tagging template according to a required data object structure;

obtaining, by the computer system, the tagging template for the unstructured document, the tagging template including the plurality of tagging elements, each tagging element having an element classification, the tagging template being obtained based on a particular ontology and a description of the unstructured document, the plurality of tagging elements of the tagging template corresponding to one or more properties of the particular ontology;

receiving, by the computer system, from a user via an interface, a plurality of selected terms corresponding to the plurality of tagging elements, wherein the term classifications of the selected terms matches the element classification of the corresponding tagging elements;

generating, by the computer system from the unstructured document based on the plurality of selected terms corresponding to the plurality of tagging elements, a data object organized according to the particular ontology, the generated data object conforming to the required data object structure;

determining a selected tagging element from the plurality of tagging elements;

prompting to the user, via the interface, suggested terms from the plurality of selected terms having a term classification matching the element classification of the selected tagging element; and

receiving, from the user, a narrative structure indicative of at least one relationship between the plurality of tagging elements of the tagging template.

9. The method of claim 8 , further comprising providing, by the computer system, to the user via the interface, a visual display of the plurality of tagging elements and the narrative structure indicating the relationship between the plurality of tagging elements.

10. The method of claim 8 , further comprising:

determining, by the computer system a selected tagging element from the plurality of tagging elements; and

providing, by the computer system, to the user, a prompt providing information about at least one characteristic of the selected tagging element.

11. The method of claim 8 , further comprising:

determining, by the computer system, a first selected tagging element from the plurality of tagging elements;

identifying, by the computer system, the first selected tagging element to the user;

receiving, by the computer system, a first selected term corresponding to the first selected tagging element according to user input;

determining, by the computer system a second selected tagging element from the plurality of tagging elements;

identifying, by the computer system, the second selected tagging element to the user; and

receiving, by the computer system, a second selected term corresponding to the second selected tagging element according to user input,

wherein the plurality of selected terms includes at least the first selected term and the second selected term.

12. The method of claim 8 , further comprising:

receiving, by the computer system, from a user, a selection of a second data object; and

storing, by the computer system, a data link between the generated data object and the selected second data object.

13. The method of claim 8 , further comprising:

receiving, by the computer system, from a second user, a second plurality of selected terms corresponding to the plurality of tagging elements, wherein the term classifications of the selected terms matches the element classification of the corresponding tagging elements;

receiving, by the computer system, from a third user, a third plurality of selected terms corresponding to the plurality of tagging elements, wherein the term classifications of the selected terms matches the element classification of the corresponding tagging elements; and

wherein generating the data object from the unstructured document further includes generating the data object according to the plurality of selected terms, the second plurality of selected terms, and the third plurality of selected terms.

14. The method of claim 8 , wherein assigning the term classification to the plurality of terms further includes assigning, by the computer system, to each of the plurality of terms at least one of a part of speech and a grammatical role.

Assignments (3)
SECURITY INTEREST Recorded Jul 3, 2022
From: PALANTIR TECHNOLOGIES INC.
To: WELLS FARGO BANK, N.A.
Reel/Frame 060572/0506 →
SECURITY INTEREST Recorded Jun 4, 2020
From: PALANTIR TECHNOLOGIES INC.
To: MORGAN STANLEY SENIOR FUNDING, INC.
Reel/Frame 052856/0817 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded May 22, 2018
From: MARC-AURELE, BRANDON; DOYLE, JOHN
To: PALANTIR TECHNOLOGIES INC.
Reel/Frame 045871/0136 →
Cited By (3)
US 12,254,263 US 12,361,045 US 12,651,018