Producing Just Enough Documentation: An Optimization Approach Applied to the Software Architecture Domain

The Software Architecture is an important asset in a software development process, which serves to share and discuss the main design concerns among the project stakeholders. The architecture must be properly documented (e.g., via a Wiki environment) to be effectively used by these stakeholders. Howe...

ver descrição completa

Detalhes bibliográficos
Autores: Diaz Pace, Jorge Andres, Villavicencio, Christian Paulo, Schiaffino, Silvia Noemi, Nicoletti, Matías Alberto, Vázquez, Hernán Ceferino
Formato: artículo
Estado:Versión publicada
Fecha de publicación:2016
País:Argentina
Recursos:Consejo Nacional de Investigaciones Científicas y Técnicas
Repositorio:CONICET Digital (CONICET)
Idioma:inglés
OAI Identifier:oai:ri.conicet.gov.ar:11336/58569
Acesso em linha:http://hdl.handle.net/11336/58569
Access Level:acceso abierto
Palavra-chave:ARCHITECTURE DOCUMENTATION MODEL
DISCRETE OPTIMIZATION
INFORMATION NEEDS
ROBUSTNESS
SENSITIVITY ANALYSIS
STAKEHOLDERS
https://purl.org/becyt/ford/1.2
https://purl.org/becyt/ford/1
Descrição
Resumo:The Software Architecture is an important asset in a software development process, which serves to share and discuss the main design concerns among the project stakeholders. The architecture must be properly documented (e.g., via a Wiki environment) to be effectively used by these stakeholders. However, the process of producing architecture documentation often fails to deliver contents that address the stakeholders’ information needs. To address the problem, we argue for a knowledge management strategy in which: (i) architecture documentation is created incrementally; and (ii) its contents are driven by a model of stakeholder preferences. In this work, we present an information optimization approach applied to the architecture documentation domain, derived from an existing documentation method called Views & Beyond. To do so, we define the Next SAD Version Problem (NSVP) and then provide tool support to assist architects in producing cost-effective documentation. Based on prior work, we perform a sensitivity analysis of the optimization model and develop a robust formulation that takes into account uncertainty in the parameter estimations for NSVP instances, thus improving the outcomes of our documentation assistant.