Using OCR to read handwritten texts in search for NFRs in Agile Software Engineering

  • Richard Rabin Maiti Berea College
  • Aleksandr Krasnov
  • Deanna Marie Wilborne
  • Frank Mitropoulos
Keywords: Agile Software Engineering, Capture Elicit Prioritize, CEP, Functional Requirements, Non-Functional Requirements, NFRs, FRs, NERV, NORMAP, Optical Character Recognition, OCR

Abstract

Non-Functional requirements (NFRs) are ignored in agile software engineering and Functional Requirements (FRs) often take center stage during agile software development due to the nature of agile software.  Research shows neglecting NFRs can be expensive. By extending the Capture Elicit Predict (CEP) methodology, to extract NFRs from handwritten text that are written by stakeholders and team members during the beginning stages of agile software development, NFRs can be identified early.  Research has shown that NFRs are helpful and compliment FRs during the beginning stages of agile software engineering.  This research shows that using NFRs from handwritten texts from requirements 3 x 5 cards can be beneficial in the agile software development process.  The OCR recognition of handwriting was 74.13%.  The accuracy of the NFRs that were captured were 92%. 

Author Biographies

Aleksandr Krasnov

Berea College - Department of Computer Science 

Deanna Marie Wilborne

Nova Southeastern University - College of Engineering & Computing 

Frank Mitropoulos

Nova Southeastern University - College of Engineering & Computing

Published
2019-07-28
Section
Articles