View source for 18450588. METHOD AND SYSTEM FOR DOCUMENT STRUCTURE BASED UNSUPERVISED LONG-FORM TECHNICAL QUESTION GENERATION simplified abstract (Tata Consultancy Services Limited)

=METHOD AND SYSTEM FOR DOCUMENT STRUCTURE BASED UNSUPERVISED LONG-FORM TECHNICAL QUESTION GENERATION=


==Organization Name==

[[:Category:Tata Consultancy Services Limited|Tata Consultancy Services Limited]]


[[Category:Tata Consultancy Services Limited]]

==Inventor(s)==

[[:Category:SUBHASISH Ghosh of Kolkata (IN)|SUBHASISH Ghosh of Kolkata (IN)]][[Category:SUBHASISH Ghosh of Kolkata (IN)]]

[[:Category:ARPITA Kundu of Kolkata (IN)|ARPITA Kundu of Kolkata (IN)]][[Category:ARPITA Kundu of Kolkata (IN)]]

[[:Category:INDRAJIT Bhattacharya of Kolkata (IN)|INDRAJIT Bhattacharya of Kolkata (IN)]][[Category:INDRAJIT Bhattacharya of Kolkata (IN)]]

[[:Category:PRATIK Saini of Noida (IN)|PRATIK Saini of Noida (IN)]][[Category:PRATIK Saini of Noida (IN)]]

[[:Category:TAPAS Nayak of Kolkata (IN)|TAPAS Nayak of Kolkata (IN)]][[Category:TAPAS Nayak of Kolkata (IN)]]

==METHOD AND SYSTEM FOR DOCUMENT STRUCTURE BASED UNSUPERVISED LONG-FORM TECHNICAL QUESTION GENERATION - A simplified explanation of the abstract==


This abstract first appeared for US patent application 18450588 titled 'METHOD AND SYSTEM FOR DOCUMENT STRUCTURE BASED UNSUPERVISED LONG-FORM TECHNICAL QUESTION GENERATION

==Simplified Explanation==

The present disclosure describes a method for generating technical questions based on the structure of a document, using Natural Language Processing techniques.

* The system first receives a textbook document and extracts PDF metadata from it using NLP.
* It then filters a variety of structures from the document based on the PDF metadata.
* Next, it obtains question templates from predefined templates using the structures, including index and Table of Contents based templates.
* The system generates long-form technical questions using the obtained question templates.
* The generated questions are evaluated using various metrics and used to improve a supervised question generation model.

== Potential Applications ==
This technology could be applied in educational settings to automatically generate practice questions for students based on textbook content. It could also be used in training programs to create assessments for employees based on training materials.

== Problems Solved ==
This technology solves the problem of manually creating technical questions for educational or training purposes. It streamlines the process by automating question generation based on document structure.

== Benefits ==
The benefits of this technology include saving time and effort in creating technical questions, ensuring consistency in question quality, and providing a more efficient way to assess knowledge based on document content.

== Potential Commercial Applications ==
Potential commercial applications of this technology include educational software platforms, training programs for businesses, and online learning platforms that can benefit from automated question generation based on document structure.

== Possible Prior Art ==
One possible prior art for this technology could be existing question generation systems that use NLP techniques, but may not specifically focus on generating technical questions based on document structure.

== Unanswered Questions ==
=== How does the system handle complex document structures with multiple levels of hierarchy? ===
The article does not delve into the specifics of how the system navigates complex document structures to generate relevant technical questions.

=== What is the accuracy rate of the generated questions compared to manually created questions? ===
The article does not provide information on the accuracy rate of the generated questions and how it compares to questions created manually.


==Original Abstract Submitted==

The present disclosure a method for document structure based unsupervised long-form technical question generation. Initially, the system receives a textbook document. Further, a PDF metadata is extracted from the textbook document using a Natural Language Processing (NLP) technique. Further, a plurality of structures from the textbook document based on the PDF metadata using an NLP based filtering technique. Further, a plurality of index based question templates and Table of Contents (TOC) based question templates are obtained from a plurality of predefined question templates using the plurality of structures. Further, the generated plurality of long-form technical questions are generated using the obtained index and TOC based question templates. The plurality of long-form technical questions are further evaluated by the system using plurality of metrics. Further, the generated plurality of long-form technical questions are used to finetune a supervised question generation model for generating optimal questions from document structure.



[[Category:G06F40/40]]
[[Category:G06F40/137]]
[[Category:G06F40/205]]
[[Category:G06Q50/20]]
[[Category:G06V30/413]]