Class diagram extraction from textual requirements using Natural language processing (NLP) techniques
Class diagram extraction from textual requirements using Natural language processing (NLP) techniques
Abstract
The automation of class generation from natural language requirements is highly challenging. This paper proposes a method and a tool to facilitate requirements analysis process and class diagram extraction from textual requirements using natural language processing NLP and Domain Ontology techniques. Requirements engineers analyze requirements manually to come out with analysis artifacts such as class diagram. The time spent on the analysis and the low quality of human analysis proved the need of automated support. A “Requirements Analysis and Class Diagram Extraction (RACE)” is a desktop instrument to assist requirements analysts and students in analyzing textual requirements, finding core concepts and its relationships, and step by step extraction of the class diagram. The evaluation of RACE system is in the process and will be conducted using two forms of evaluation, student and expert evaluation.
.
Keywords: Natural language processing (NLP), Domain Ontology, UML Class Diagram.
1. Introduction
The common way to express requirements is with large volumes of text [1] which can be referred to as natural language (NL) requirements. NL requirements are typically coming from a pool of natural language statements which are gathered from interview excerpts, documents and notes. Due to the inherent ambiguity of natural language, it is often difficult to prove properties on natural language requirements [2]. For this reason, Informal natural language requirements are better to be expressed as formal representations.
Object-Oriented Analysis and Design (OOAD) has become a popular approach for software development since the 1990’s [1]. UML class diagrams are the main core of OO analysis and design systems where most other models are derived from [3].
Natural language processing (NLP) is recognized as a general assistance in analyzing
Abstract
The automation of class generation from natural language requirements is highly challenging. This paper proposes a method and a tool to facilitate requirements analysis process and class diagram extraction from textual requirements using natural language processing NLP and Domain Ontology techniques. Requirements engineers analyze requirements manually to come out with analysis artifacts such as class diagram. The time spent on the analysis and the low quality of human analysis proved the need of automated support. A “Requirements Analysis and Class Diagram Extraction (RACE)” is a desktop instrument to assist requirements analysts and students in analyzing textual requirements, finding core concepts and its relationships, and step by step extraction of the class diagram. The evaluation of RACE system is in the process and will be conducted using two forms of evaluation, student and expert evaluation.
.
Keywords: Natural language processing (NLP), Domain Ontology, UML Class Diagram.
1. Introduction
The common way to express requirements is with large volumes of text [1] which can be referred to as natural language (NL) requirements. NL requirements are typically coming from a pool of natural language statements which are gathered from interview excerpts, documents and notes. Due to the inherent ambiguity of natural language, it is often difficult to prove properties on natural language requirements [2]. For this reason, Informal natural language requirements are better to be expressed as formal representations.
Object-Oriented Analysis and Design (OOAD) has become a popular approach for software development since the 1990’s [1]. UML class diagrams are the main core of OO analysis and design systems where most other models are derived from [3].
Natural language processing (NLP) is recognized as a general assistance in analyzing
References: [1] Booch, G. (1994). Object-Oriented Analysis and Design with Applications, 2nd Ed., Benjamin Cummings. [2] Ambriola, V [5] Elizabeth D. Liddy & Jennifer H. Liddy, 2001, “An NLP Approach for Improving Access to Statistical Information for the Masses”. [6] Gobinda G. Chowdhury , 2001, Natural Language Processing. [9] Xiaohua Zhou and Nan Zhou, 2004, Auto-generation of Class Diagram from Free-text Functional Specifications and Domain Ontology [10] L [16] Jawad Makki, Anne-Marie Alquier, and Violaine Prince, 2008 Ontology Population via NLP techniques in Risk Management, ICSWE: Fifth International Conference on Semantic Web Engineering, Heidelberg, Germany , v.1