Automated formal specification generation and refinement from requirement documents

Cabral, Gustavo Augusto Porto Sereno; Sampaio, Augusto

doi:10.1590/s0104-65002008000100008

Cited by 1 publication

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, most of these transformation methods largely depend on the Natural Language Processing technique which is still facing the problem of the great amount of ambiguity issues at every linguistic level of natural language [5]. Although Controlled Natural Languages are proposed to reduce these ambiguities to some extent by restricting the grammar and vocabulary [6], [7], they need to be studied and their completeness is hard to guarantee. There are also some intermediate languages being invented [8], [9], but developers still need to deal with details concerned with formal notation when reaching the bottom level of formal specifications.…”

Section: Introductionmentioning

confidence: 99%

Computer-Aided Formalization of Requirements Based on Patterns

Wang

Liu

2014

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYFormalizing requirements in formal specifications is an effective way to deepen the understanding of the envisioned system and reduce ambiguities in the original requirements. However, it requires mathematical sophistication and considerable experience in using formal notations, which remains a challenge to many practitioners. To handle this challenge, this paper describes a pattern-based approach to facilitate the formalization of requirements. In this approach, a pattern system is predefined to guide requirements formalization where each pattern provides a specific solution for formalizing one kind of function into a formal expression. All of the patterns are classified and organized into a hierarchical structure according to the functions they can be used to formalize. The distinct characteristic of our approach is that all of the patterns are stored on computer as knowledge for creating effective guidance to facilitate the developer in requirements formalization; they are "understood" only by the computer but transparent to the developer. We also describe a prototype tool that supports the approach. It adopts Hierarchical Finite State Machine (HFSM) to represent the pattern knowledge and implements an algorithm for applying it to assist requirements formalization. Two experiments on the tool are presented to demonstrate the effectiveness of the approach.

show abstract