Michal Kosinár

Building process definition with ontology background

Documented software processes and their assessments are the basics of modern software development. Currently the semantic web, knowledge bases and knowledge management have many applications. Yet, applications to support software processes (and business processes in general) are rather neglected. In this paper we focus on applying a knowledge layer into software processes and on the design of such a knowledge base. After a brief description of some classical fundamental approaches to software processes and knowledge bases, we propose an improvement based on the application of a machine readable knowledge base. In particular, we aim at optimizing and improving software process development using knowledge bases created with the aim to formally describe software process development.

Know How and Know What for Software Processes

Formal specification of a software process, as well as its optimal design, is a fundamental landmark and tenet that any successful software company must follow. Recent trends can be characterized as a knowledge-base support of the software-process development, standardization and improvement. To this end we create semantic annotations (ontologies) of processes which should serve as a stable unifying core of the software-process development. However, when doing so, we meet the problem how to transform various forms of tacit, implicit knowledge into an explicit knowledge specification that is logically tractable and machine readable. In this paper we focus on the transformation of informal tacit knowledge about a software process (or any part of the process) to the formal knowledge specification that can be used for building machine readable knowledge bases. In particular, we aim at optimizing and improving software-process development using knowledge bases which are created to the purpose of a formal description of the software-process development.

Simulations of Agile Software Processes for Healthcare Information Systems Development Based on Machine Learning Methods

Abstract Requirements on software products are becoming more and more complicated and software systems of today are characterized by increasing complexity and size. Therefore, software systems can no longer be developed feasibly without the processes supported by appropriate methods. We propose a method for configuration and modification of software processes in companies based on gathered knowledge and our approach allows to support and optimize the processes with formal methods of modelling and machine-learning based simulations.