Construção e avaliação empírica de um catálogo de padrões de requisitos para sistemas de registro eletrônico em saúde
Nenhuma Miniatura disponível
Data
2022-03-31
Autores
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Federal de Goiás
Resumo
[Context] Requirements pattern catalogs (CPR) are a useful reuse approach for dealing
with incomplete, incorrect, and omitted requirements. The main benefits of using catalogs
are: time savings and higher quality requirements. [Problem] Despite these advantages,
there is a scarcity of supported domains and of empirical studies evaluating the use of
these catalogs. [Objective] - Develop a CPR for Electronic Health Record Systems (SEHR) and evaluate it empirically considering: (a) support for requirements elicitation
and specification and (b) the perception of its usefulness and ease of use. [Methods] -
The requirements modeled in the CPR are based on a Certification Manual for S-RES
marketed in Brazil. The structure and grammar of the CPR is based on a metamodel
called Software Pattern Metamodel (SoPaMM), and the construction of the CPR took
place using the Terminal Model Editor (TMed) tool. In two case studies, the quality of
requirements documents built with and without the use of CPR by undergraduate students
was evaluated. A survey was also applied to collect students perception of the usefulness
and ease of use of CPR. [Results] - In terms of completeness and scope, the catalog
helped to increase the quality of the Requirements especification document (DER) by
approximately 50%. The catalog also made it possible to eliminate inconsistencies in
case of reuse. In the perception analysis, 98% of the students stated that the use of the
catalog enabled faster elicitation and specification. [Conclusions] - It was identified that
the approach helps to improve the quality of the DER and allows a perception of faster
elicitation and specification.
Descrição
Palavras-chave
Citação
MARTINS, M. C. Construção e Avaliação Empírica de um Catálogo de Padrões de Requisitos para Sistemas de Registro Eletrônico em Saúde. 2022. 107 f. Dissertação (Mestrado em Ciência da Computação), Universidade Federal de Goiás, Goiânia, 2022.