An Intelligent Framework to Generate Use Case Diagrams and Class Diagrams from Requirements Documents

Abstract

Use case and class diagramming are essential requirements engineering techniques that play a pivotal role in modeling software specifications and facilitating the software development process. However, software requirements are often expressed in Natural Language (NL), which can be ambiguous, noisy, immeasurable, and open to interpretation. This research addresses these challenges by automatically extracting the required elements to generate use case and class diagrams from software requirements documents written in natural language. Accordingly, an automated framework is proposed based on Natural Language Processing (NLP) techniques—such as tokenization and part-of-speech tagging—to parse the software requirements syntactically using a set of heuristic rules. These rules facilitate the extraction of actors, use cases, entities, relationships, and attributes required for generating the corresponding diagrams. Furthermore, to enhance the framework’s performance, the k-nearest neighbor (k-NN) algorithm is employed to predict previously processed requirements and reduce redundant computation. The framework’s effectiveness was evaluated using two performance metrics: recall and precision. Experimental results show that the proposed approach achieves an average recall of 96% and an average precision of 92%, confirming its robustness and reliability.

Keywords