Please use this identifier to cite or link to this item: https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4634
Title: AI Based UML Diagrams Generator
Authors: Arachchi, K. A. D Oshada Kasun Kiringoda
Keywords: UML Diagram
Use case Diagram
Class Diagram
Natural Language Processing,
Machine Learning
Classification
RNN
sequence-to-sequence RNN
Plant UML
Issue Date: 19-Aug-2022
Abstract: Unified Modeling Language (UML) diagrams are very useful to represent the business requirement of any proposed system and help to design a system from the end user's point of view. In the software analysis process, UML diagrams are drawn separately after gathering requirement, and the time wastage of drawing these diagrams is high with using current drawing tools due to the complexity of the business situation or the technical capabilities of the UML diagrams. Automated UML Diagram generation tools necessity can be identified with amount of time spent on requirement analysis and low quality of human analysis. The objective of this project is to provide an approach to generating UML use case and class diagrams from the functional requirement texts using natural language processing and machine learning. Finding the key terms for class or use case diagram is reflected a classification task in machine learning. The stories narrate different attributes of a diagram and the task is to identify key terms for respective attributes. The use case diagram and class diagram have different set of attributes. classification model can be designed and pre-process the data for extracting the key terms such as actors, use case and classes from the requirement text. Furthermore, other text features like position and distance of text can be integrated to improve accuracy. To Identify Use case and class relationship sequence to sequence RNN model has introduced and it extracted key phrase or relation phrase among text, which also outperformed the conventional technique significantly. Improved strategies were proposed on this approach by substituting a sequence-to-sequence RNNs with conventional techniques. The proposed system is capable of providing solutions to generating usecase and class diagram from the functional requirement text with using NLP and ML techniques. Diagram generating has used Plant UML tool with its plan text language where user interaction also enabled. Developed prototype is able to identify important actors, use cases and its relations with relationship type for use case diagrams. And also for class diagrams important classes and it’s relationships can identify. If this project is domain specific most of element would have been identified. The developed final model has identified and generated the use case diagram and class diagram up to a considerable extent according to the given requirement text which gives a good idea about the business scenario.
URI: https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4634
Appears in Collections:2021

Files in This Item:
File Description SizeFormat 
2018 MCS 047.pdf3.95 MBAdobe PDFView/Open


Items in UCSC Digital Library are protected by copyright, with all rights reserved, unless otherwise indicated.