Thorsten Merten

Towards a new understanding of small and medium sized enterprises in requirements engineering research

[Context and motivation] Almost worldwide the software industry mainly consists of small and medium software enterprises. From a requirements engineering perspective these companies are poorly researched. [Problem] Though RE research is discovering SMEs as an interesting field, it is difficult to categorize and distinguish these companies sufficiently. This leads to a) weakly classified results of observational studies as well as field studies and empirical research and b) insufficient mappings between methodical improvements and the companies they can be applied to. Therefore, it is hard for researchers and enterprises to adopt RE state of the art to an enterprises environment. [Principal ideas] After defining the problem, initial ideas for attributes classifying SMEs are presented and a way for improving and clustering these attributes is shown. [Contribution] This paper raises an important problem statement for RE research and shows an initial way towards solving this problem.

Classifying unstructured data into natural language text and technical information

Software repository data, for example in issue tracking systems, include natural language text and technical information, which includes anything from log files via code snippets to stack traces. However, data mining is often only interested in one of the two types e.g. in natural language text when looking at text mining. Regardless of which type is being investigated, any techniques used have to deal with noise caused by fragments of the other type i.e. methods interested in natural language have to deal with technical fragments and vice versa. This paper proposes an approach to classify unstructured data, e.g. development documents, into natural language text and technical information using a mixture of text heuristics and agglomerative hierarchical clustering. The approach was evaluated using 225 manually annotated text passages from developer emails and issue tracker data. Using white space tokenization as a basis, the overall precision of the approach is 0.84 and the recall is 0.85.

Using RE knowledge to assist automatically during requirement specification

In a two semester software engineering (SE) course at Bonn-Rhine-Sieg University students have the opportunity to actually elicit, analyze and document requirements as well as design and develop a correspondent software product in teams of approximately four. The students have to use an issue tracking software in combination with a Requirements Engineering (RE) tool to document and plan their work. Though the course starts with RE theory from elicitation via documentation and traceability, we found that the students find it difficult to combine different RE artifact types and to develop useful traces between them. In this paper we present an approach to provide feedback and give pro-active advice inside an RE tool, while the specification is created. To derive this feedback we use a knowledge base containing rules and best practices to create a requirements specification. An assistance system applies these rules to guide the user in different situations, beginning with an empty specification up to the implementation of various RE artifact types and traces between them. This paper presents the status of our knowledge-based feedback mechanism and possible extensions. In order to get primary indicators for the value of this approach we did experiments and workshops with eight students who worked with the same tool with and without the feedback system.

Tooling zur schrittweisen methodischen Fundierung des RE in KMU

Strukturiertes und methodisch fundiertes RE kann in kleinen und mittelstandischen Software-Unternehmen (im Folgenden KMU) aus mehreren Grunden eingefuhrt werden. Beispiele hierfur sind Firmenwachstum, Kundenvorgaben, Fluktuation oder intrinsische Motivation des KMU. Wachsende Firmen konnen ab einer gewissen Mitarbeiterzahl von ca. 15-20 Personen nicht mehr auf eine rein verbale Kommuninkation von Anforderungen zuruckgreifen und mussen explizites RE bzw. entsprechende Prozesse einfuhren [2]. Eine Forderung von Stakeholdern, bspw. seitens des Kunden, nach bestimmten Vorgehensweisen oder Dokumenten (wie Pflichtenoder Lastenheft) konnen KMU motivieren, die Erstellung dieser Dokumente in einem strukturierten RE-Prozess zu betreiben. Die Gefahr beim plotzlichen Ausscheiden eines Team-Mitglieds (auch bekannt als Truck-Faktor [8, S. 41]) stellt eine weitere mogliche intrinsische Motivation von KMU dar, Anforderungswissen oder sonstiges Wissen zu dokumentieren.