- Patent Title: Recommending form field augmentation based upon unstructured data
-
Application No.: US15013845Application Date: 2016-02-02
-
Publication No.: US10248639B2Publication Date: 2019-04-02
- Inventor: Pankaj S. Dayama , Varun R. Embar , Vinayaka Pandit
- Applicant: International Business Machines Corporation
- Applicant Address: US NY Armonk
- Assignee: INTERNATIONAL BUSINESS MAHCINES CORPORATION
- Current Assignee: INTERNATIONAL BUSINESS MAHCINES CORPORATION
- Current Assignee Address: US NY Armonk
- Agency: Ference & Associates LLC
- Main IPC: G06F17/24
- IPC: G06F17/24 ; G06F17/30

Abstract:
One embodiment provides a method for recommending a structured field for a form from unstructured text data, the method including: utilizing at least one processor to execute computer code that performs the steps of: obtaining text data from at least one unstructured field, wherein the at least one unstructured field is contained within a completed form generated from a template form; identifying at least one topic associated with the text data; generating a model, wherein the model analyzes use of the least one topic as a structured field; determining, using the model, whether the accuracy of the template form has increased based upon use of the at least one topic as a structured field; and recommending, based upon the determining, at least one modification for a structured field for the template form, wherein the at least one structured field is associated with the at least topic.
Public/Granted literature
- US20170220541A1 RECOMMENDING FORM FIELD AUGMENTATION BASED UPON UNSTRUCTURED DATA Public/Granted day:2017-08-03
Information query