Systems and methods for locating, identifying and mapping electronic form fields

• US 10,133,718 B2
• Filed: 07/20/2015
• Issued: 11/20/2018
• Est. Priority Date: 07/18/2014
• Status: Active Grant

• Current Assignee: UBIQ Security Incorporated
• Original Assignee: FHOOSH, Inc.
• Inventors: Iasi, Anthony, Eigner, Linda, Eigner, William, Kahle, Charles, Tobias, Eric
• Primary Examiner(s): Bloomquist, Keith D
• Application Number:
  US14/804,314
• Publication Number:
  US 20160019197A1
• Time in Patent Office: 1,219 Days
• CPC Class Codes: G06F 40/174 Form filling; Merging
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

Systems and methods for locating, identifying, mapping and completing electronic form fields are provided herein. A mapping engine is configured to identify form fields using a variety of similar field names through one or more algorithms configured to identify and match similar field names and combinations of field names. A form field mapping and identification engine identifies a form category using a machine learning classification algorithm, then determines and maps form labels to form fields using seeded values and optical scanning in order to produce a human readable label for each form field. The field labels are used to generate a set of terms for each form field that are used to identify content to be filled in the form with a high degree of accuracy. Additional embodiments are directed toward locating form fields in an electronic form known as a formless form.

Claims

1. A method of matching and completing form fields in an electronic form, the method comprising the steps of:

receiving at least one field name of at least one form field on the electronic form, the at least one field name comprising multiple terms, the electronic form comprising a plurality of form fields, a plurality of field names associated with the plurality of form fields, and a plurality of readable labels;

decomposing the electronic form into a hierarchical structure in descending order of form category type, section headers, and subsections, wherein the section headers and subsections each comprise one or more form fields of the plurality of form fields;

scanning the plurality of readable labels for keywords in the electronic form;

applying a probabilistic algorithm to the keywords to identify a form category type for the electronic form;

building associations between the plurality of readable labels and the plurality of form fields to determine a readable label of the plurality of readable labels for the at least one field name associated with the at least one form field;

extracting contextual information from the electronic form based on at least one of a section header corresponding to the at least one form field and a subsection corresponding to the at least one form field;

composing a set of terms for the at least one form field based on the identified form category type, the determined readable label, and the contextual information, wherein the set of terms is not the same as the readable label;

performing a best-fit search of a database of field names using the set of terms to identify a best-fit field name of the database of field names for the at least one form field; and

transmitting a field value corresponding to the best-fit field name for completing the at least one form field of the electronic form.

2. The method of claim 1, wherein the field name of the form field is different than the best-fit field name in the database of field names and corresponding field values.

3. The method of claim 1, wherein the best-fit search utilizes an algorithm to identify the best-fit field name.

4. The method of claim 1, wherein the best-fit search improves over time based on continuously updated knowledge of user entries of field values into form fields.

5. The method of claim 1, wherein the multiple terms of the field name are determined based on a hierarchical structure of the electronic form.

6. The method of claim 1, wherein the field names are renamed using human readable labels based on the association between the readable labels and the at least one form field.

7. The method of claim 1, further comprising:

identifying a subset of form fields relevant to the electronic form based on the form category type.

8. The method of claim 1, wherein the form category type is identified based in part on a frequency of occurrence of one or more keywords.

9. The method of claim 1, wherein building associations comprises:

generating a first mapping based on associating a sequential set of seeded values with a subset of the plurality of field names corresponding to a subset of the plurality of form fields;

generating a second mapping based on associating the sequential set of seeded values with a subset of plurality of readable labels corresponding to the subset of the plurality of form fields; and

associating each of the subset of the plurality of readable labels to each of the subset of the plurality of field names based on the first and second mappings.

10. A system for matching and completing form fields in an electronic form, the system comprising:

a client device accessing the electronic form comprising a plurality of form fields, a plurality of field names associated with the plurality of form fields, and a plurality of readable labels;

a server in communication with the client device which receives a field name of a form field on the accessed electronic form, the field name comprising a multiple terms;

a content database in communication with the server, wherein the content database stores a plurality of pre-defined field names and corresponding field values;

wherein the server is configured to;

decompose the electronic form into a hierarchical structure in descending order of form category type, section headers, and subsections, wherein the section headers and subsections each comprise one or more form fields of the plurality of form fields;

scan the plurality of readable labels for keywords in the electronic form;

apply a probabilistic algorithm to the keywords to identify a form category type for the electronic form;

build associations between the plurality of readable labels and the plurality of form fields to determine a readable label of the plurality of readable labels for the received field name of the form field;

extract contextual information from the electronic form based on at least one of a section header corresponding to the received form field and a subsection corresponding to the received form field;

compose a set of terms for the at least one form field based on the identified form category type, the determined readable label, and the contextual information, wherein the set of terms is not the same as the readable label; and

perform a best-fit search of the field names in the content database using the set of terms to identify a best-fit field name from the pre-defined field names of the content database and a field value corresponding to for the best-fit field name for the received form field, which is then transmitted to the client device for filling into the received form field.

11. The system of claim 10, wherein the field name of the form field is different than the best-fit field name in the content database of pre-defined field names and corresponding field values.

12. The system of claim 10, wherein the best-fit search utilizes an algorithm to identify the best-fit field name.

13. The system of claim 10, wherein the best-fit search improves over time based on continuously updated knowledge of user entries of field values into form fields.

14. The system of claim 10, wherein the multiple terms of the field name are determined based on a hierarchical structure of the electronic form.

15. The system of claim 10, wherein the field names are renamed using human readable labels based on the association between the readable labels and the at least one form field.

16. The system of claim 10, wherein the server is configured to:

identify a subset of form fields relevant to the electronic form based on the form category type.

17. A method of identifying and labeling fields in an electronic form, the method comprising the steps of:

decomposing the electronic form into a hierarchical structure in descending order of form category type, section headers, and subsections, the electronic form comprising a plurality of form fields, a plurality of field names associated with the plurality of form fields, and a plurality of readable labels, wherein the section headers and subsections each comprise one or more form fields of the plurality of form fields;

determining a form category of the electronic form by applying a probabilistic algorithm using the hierarchical structure of the electronic form and keywords extracted from the electronic form;

associating the plurality of readable labels with the plurality of form fields to determine a readable label of the plurality of readable labels for at least one field name associated with at least one form field;

extracting contextual information from the form using an optical scanner based on at least one of a section header corresponding to the at least one form field and a subsection corresponding to the at least one form field;

composing at least one term for each form field in the electronic form based on the determined form category, the determined readable label, and the contextual information; and

assigning the composed at least one term for each of the form fields to each of the field name.