Method and system for determining action items using neural networks from knowledge base for execution of operations

Patent metadata

Field	Value
Publication number	US-10318636-B2
Application number	US-201615372020-A
Country	US
Kind code	B2
Filing date	Dec 7, 2016
Priority date	Oct 30, 2016
Publication date	Jun 11, 2019
Grant date	Jun 11, 2019

Abstract

Systems and methods for determining action items from knowledge base for execution of operation. The system receives instructions, present in a knowledge base, which are required to execute one or more operations. Thereafter, the system parses the instructions into one or more sentences and assigns a POS tag for each word in the one or more sentences. Further, the system assigns a predefined class for each of the POS tagged word. Based on the predefined class, the system determines the action items. The action item comprises one or more actions and one or more components on which the one or more actions are to be performed. The present disclosure enables automated systems to easily execute one or more operation based on the action items thereby reducing the delay in performance of the automated system due to complexity in interpreting the instructions.

First claim

What is claimed is:

1. A method for determining action items from a knowledge base for execution of operations, the method implemented by a computing device and comprising:
receiving one or more instructions present in a knowledge base, wherein the one or more instructions are required to execute one or more operations;
automated parsing of each of the one or more instructions into one or more sentences;
automated tagging each of one or more words in the one or more sentences with one of a plurality of Part of Speech (POS) tags;
assigning with a neural network classifier one of a plurality of classes for each of the POS tagged words in the one or more sentences based on one or more predefined parameters, wherein the plurality of classes are all different from the plurality of POS tags and the plurality of classes comprise at least an identified action and an identified component;
determining one or more computer system executable action items in each of the one or more sentences based on the assigned one of the plurality of classes for each of the POS tagged words for executing the one or more operations; and
providing the determined one or more computer system executable action items to an automated system to execute.

2. The method as claimed in claim 1 , wherein the one or more predefined parameters comprise the POS tag for a target one of the words in each of the one or more sentences, the POS tag for two of the words prior to the target word, or a word embedding of the target word.

3. The method as claimed in claim 1 , further comprising selectively generating one or more noun phrases for the one or more sentences comprising the POS tagged words prior to the assigning the one of the plurality of classes to each of the POS tagged words.

4. The method as claimed in claim 1 , wherein each of the one or more instructions are in a natural text format.

5. The method as claimed in claim 1 , wherein the assigned one of the plurality of classes is one of an action, a begin component, or an inside component.

6. The method as claimed in claim 1 , wherein the action items comprise one or more actions to be performed and corresponding one or more components on which the one or more actions are to be performed.

7. An action item determination computing device comprising a memory comprising programmed instructions stored thereon, the memory coupled to one or more processors that are configured to be capable of executing the stored programmed instructions to:
receive one or more instructions present in a knowledge base, wherein the one or more instructions are required to execute one or more operations;
automated parse of each of the one or more instructions into one or more sentences;
automated tag each of one or more words in the one or more sentences with one of a plurality of Part of Speech (POS) tags;
assign with a neural network classifier one of a plurality of classes for each of the POS tagged words in the one or more sentences based on one or more predefined parameters, wherein the plurality of classes are all different from the plurality of POS tags and the plurality of classes comprise at least an identified action and an identified component;
determine one or more computer system executable action items in each of the one or more sentences based on the assigned one of the plurality of classes for each of the POS tagged words for executing the one or more operations; and
provide the determined one or more actions computer system executable action items to an automated system to execute.

8. The action item determination computing device as claimed in claim 7 , wherein the one or more predefined parameters comprise the POS tag for a target one of the words in each of the one or more sentences, the POS tag for two of the words prior to the target word, or a word embedding of the target word.

9. The action item determination computing device as claimed in claim 7 , wherein the one or more processors are further configured to be capable of executing the stored programmed instructions to selectively generate one or more noun phrases for the one or more sentences comprising the POS tagged words prior to the assigning the predefined class one of the plurality of classes to each of the POS tagged words.

10. The action item determination computing device as claimed in claim 7 , wherein each of the one or more instructions are in a natural text format.

11. The action item determination computing device as claimed in claim 7 , wherein the assigned one of the plurality of classes is one of an action, a begin component, or an inside component.

12. The action item determination computing device as claimed in claim 7 , wherein the action items comprise one or more actions to be performed and corresponding one or more components on which the one or more actions are to be performed.

13. A non-transitory computer readable medium having stored thereon instructions for determining action items from a knowledge base for execution of operations that, when executed by one or more processors, cause the one or more processors to:
receive one or more instructions present in a knowledge base, wherein the one or more instructions are required to execute one or more operations;
automated parse of each of the one or more instructions into one or more sentences;
automated tag each of one or more words in the one or more sentences with one of a plurality of Part of Speech (POS) tags;
assign with a neural network classifier one of a plurality of classes for each of the POS tagged words in the one or more sentences based on one or more predefined parameters, wherein the plurality of classes are all different from the plurality of POS tags and the plurality of classes comprise at least an identified action and an identified component;
determine one or more computer system executable action items in each of the one or more sentences based on the assigned one of the plurality of classes for each of the POS tagged words for executing the one or more operations; and
provide the determined one or more computer system executable action items to an automated system to execute.

14. The non-transitory computer readable medium as claimed in claim 13 , wherein the one or more predefined parameters comprise the POS tag for a target one of the words in each of the one or more sentences, the POS tag for two of the words prior to the target word, or a word embedding of the target word.

15. The non-transitory computer readable medium as claimed in claim 13 , wherein the programmed instructions, when executed by the one or more processors, further cause the one or more processors to selectively generate one or more noun phrases for the one or more sentences comprising the POS tagged words prior to the assigning the one of the plurality of classes to each of the POS tagged words.

16. The non-transitory computer readable medium as claimed in claim 13 , wherein each of the one or more instructions are in a natural text format.

17. The non-transitory computer readable medium as claimed in claim 13 , wherein the assigned one of the plurality of classes is one of an action, a begin component, or an inside component.

18. The non-transitory computer readable medium as claimed in claim 13 , wherein the action items comprise one or more actions to be performed and corresponding one or more components on which the one or more actions are to be performed.

Assignees

• Wipro Ltd

Inventors

• Chatterjee Arindam
• Chaudhuri Debanjan
• KOMPELLA Anasuya Devi

Classifications

• G06F40/117

  Tagging; Marking up (details of markup languages G06F40/143); Designating a block; Setting of attributes (style sheets, e.g. eXtensible Stylesheet Language Transformation [XSLT], G06F40/154) · CPC title

• G06F40/30 Primary

  Semantic analysis · CPC title

• G06F40/205

  Parsing · CPC title

• G06F40/253

  Grammatical analysis; Style critique · CPC title

• G06F40/211

  Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars · CPC title