Vladimir A. Shekhovtsov

Towards Conceptualizing Quality-Related Stakeholder Interactions in Software Development

The paper addresses the issue of organizing quality-related interaction between business stakeholders and software developers relying on established common vocabulary. It establishes a conceptual representation for the process of such interaction. This conceptualization is based on a set of notions representing software quality and its particular incarnations; they are used to define the activities of the interaction process. The process is conceptualized on two levels: a coarse-grained level defining the set of generic activities and the conditions of launching these activities and a fine-grained level describing particular interaction steps in detail. The conceptualization is expected to be shared as a part of upcoming ontology of stakeholder quality perception and assessment.

Harmonizing the Quality View of Stakeholders

Abstract Due to the nature of software development processes, different parties (users, developers, business actors, etc.) have different views on the quality a system under development (SUD) is intended to have. Consequently, these views have to be harmonized. This chapter summarizes the related state-of-the-art and presents a conceptualization of quality view harmonization. In particular, we will deal with and conceptualize the following dimensions: the SUD quality itself, the activities of the harmonization process including the interaction of the parties involved while gathering quality requirements, the knowledge about these parties and the differences in their perception of quality, and the amount of knowledge about quality-related issues. The conceptualized knowledge of these dimensions forms the basis of an ontology for supporting the interaction process. The structure of the chapter is defined along these dimensions: For every area of knowledge we will propose a set of concepts and align the state-of-the-art approaches to them. The concepts presented here are partly a result of the ongoing project Quality Aware Software Engineering, which is funded by the Austrian Research Promotion Agency (FFG).

HCM-L: Domain Specific Modeling for Active and Assisted Living

Modeling and modeling methods are crucial for information systems engineering but are seldom seamlessly integrated into all phases of development and operation: Practitioners challenge the benefits of modeling and complain about the confusing variety of concepts with overlapping semantics, symbols and syntactic rules of today’s standardized, “universal” modeling languages. Therefore, domain-specific modeling languages (DSMLs) are gaining increasing popularity: they are lean and convenient, support the productivity of modeling, and help to increase model quality and comprehensibility. There are, however, few approaches to embedding a DSML into a domain-specific modeling method (DSMM) that provides guidelines about how to use a given DSML and to evaluate related models. This chapter aims to make a contribution towards filling that gap by discussing, as an example and proof of concept, a domain-specific modeling method for the human cognitive modeling language HCM-L, a DSML for the domain of active and assisted living. As a modeling language without tool support has no chance to be used in practice, we are conducting that discussion on the basis of HCM-L modeler, a tool that was implemented using the metamodeling platform ADOxx and can be accessed via OMiLAB, the Open Models Laboratory for modeling method engineering. HCM-L modeler is component of an ambient assistance system for supporting elder persons in mastering their daily life activities.

Towards Managing Understandability of Quality-Related Information in Software Development Processes

Establishing common understanding between the parties in the software process is important for dealing with quality of the prospective software. This process is difficult to organize because the parties (especially, developers and business stakeholders) perceive quality based on different world views. To address this problem, we aim at a solution for managing understandability of quality-related information in the software process. This solution provides the set of understandability assessment activities (aimed at diagnosing problems with communicated terms not belonging to the view of the target party) and understandability improvement activities (aimed at resolving these problems by translating problematic terms between world views and providing necessary explanations). These activities are supported by a modular ontology incorporating available quality-related knowledge; particular configuration of the ontology modules describes the quality view of the involved party. The proposed solution is expected to reduce the time and effort for establishing a communication basis while discussing software quality, thus cutting costs and strengthening the mutual trust of the parties.

Facilitating Effective Stakeholder Communication in Software Development Processes

Effective communication in software development is impaired when parties perceive communicated information differently. To address this problem, the project QuASE has been established. It aims at a solution that supports understandability and reusability of communicated information as well as the quality of decisions based on such information. In this paper, we focus on the architectural aspects of the QuASE system and on its knowledge base which consists of two ontologies: a site ontology defining the site-specific communication environment, and a “quality ontology” that incorporates all knowledge necessary for supporting communication. We describe the overall architecture of the system, introduce the ontologies as well as their interplay, and outline the approach for gathering knowledge necessary to form the QuASE site ontology.

Stakeholder involvement into quality definition and evaluation for service-oriented systems

The paper addresses the matter of quality in the software process for service-oriented systems. We argue for the need of involving the users/stakeholders into the specification and evaluation of quality (requirements) and we develop means for supporting such an involvement. For this purpose we introduce classifications of user and quality types and as a basis for the characterization of evaluation cases.

QuASE: A Tool Supported Approach to Facilitating Quality-Related Communication in Software Development

The paper describes the current state of the ongoing project QuASE. This project aims at facilitating quality-related communication in software development by the following means: a communication platform providing view harmonizing mechanisms for the different parties involved in a software process, decision making support related to quality issues, reuse of experience from former communication, prediction of quality-related experience. We describe usage scenarios for this approach, and outline its core concepts as well as the current progress of its implementation.

Acquiring Empirical Knowledge to Support Intelligent Analysis of Quality-Related Issues in Software Development

The paper describes the current state of the initial part of the ongoing project aimed at the intelligent support for dealing with quality-related information in the software process. We describe the empirical qualitative studies aimed at acquiring the knowledge on stakeholder perception of quality and quality-related stakeholder interactions. This knowledge is supposed to be incorporated into a common ontology serving as a conceptual foundation for the prospective intelligent issue monitoring and analysis system based on the semantic repository. This system can be used to support prediction of the quality-related behavior of the stakeholders and facilitate reuse of the relevant knowledge.

Model Centered Architecture

This paper advocates a rigorous model focused paradigm of information system development and use. We introduce the concept of “Model Centered Architecture” that sees an information system to be a compound of various networked models, each of which is formed with the means of a Domain Specific Modeling Language. This languages are tailored to the particular circumstances of the respective system aspect. I.e., from a MOF perspective, MCA focuses on the MOF levels M2 (definitions of the DSMLs to be used for the specification of the system and it’s contexts), M1 (Specification of all System and Data Components using the DSMLs) and M0 (the instances, i.e. models of concrete objects, functions and processes). The transformation of M0 citizens to the respective implementation concepts (Structure → Data, Function→ Program, Process→ Workflow) is delegated to mapping functions defined on M2, restricted on M1 to the particular schemata (in the sense of mappings between the respective sets of schema instances), and instantiated on M0 for the concrete instances. The paper shows how such model centered approach may be applied in practice using two real development projects as running examples.

Implementing tool support for analyzing stakeholder communications in software development

The problem with communicated information originated from stakeholders participating in software development processes is related to the fact that such information is not always available for analysis and reuse. As a rule, it is collected in project repositories (such as issue or bug databases) and only consulted “as-is”: such activities do not always satisfy the analytical needs of the software companies. The goal of a tool-supported QuASE approach presented in this paper is to allow for different kinds of analysis of the available communicated information, in particular, we support reuse of such information, prediction of the characteristics of the elements of the communication environment and the communicated information itself, issuing targeted recommendations and supporting selecting the preferred alternatives for decisions related to such information or to the elements of the communication environment. We describe the motivation for our approach, supported analytical scenarios, modeling and ontological support for analytical concepts, and the specifics of the implementation of the analysis component of the end-user software tool.