Publications of the institutes and centres - University of Applied Sciences of the Grisons

Abstract: Software archives are a rich source of information about the development process of a software system. Using data mining techniques rules can be extracted from these archives. Rules can either be association rules which mean that software items have been changed together with a certain probability or sequence rules which mean items have been changed one after the other. The change frequency (support) as well as the change probability (confidence) of a given rule are both metrics used for visualizing the strength of a rule. EPOSee is a tool designed to interactively explore these kinds of rules. To this end we extended standard visualization techniques for association and sequence rules to also show the hierarchical order of items. Clusters and outliers in the resulting visualizations provide interesting insights into the relation between the temporal development of a system and its static structure.

Abstract: Software archives contain historical information about the development process of a software system. Using data mining techniques rules can be extracted from these archives. In this paper we discuss how standard visualization techniques can be applied to interactively explore these rules. To this end we extended the standard visualization techniques for association rules and sequence rules to also show the hierarchical order of items. Clusters and outliers in the resulting visualizations provide interesting insights into the relation between the temporal development of a system and its static structure. As an example we look at the large software archive of the MOZILLA open source project. Finally we discuss what kind of regularities and anomalies we found and how these can then be leveraged to support software engineers.

Abstract: With the last four years four large student teaching projects have taken place. During these projects our approach to apply Petri nets in agent-oriented software engineering (AOSE) production has been sharpened. Within the projects different techniques, methods and tools as well as their conceptual and technical background have been developed and tried out. In this talk we will explain several of the underlying and hidden ideas, concepts and results of our work.

DOI: https://doi.org/10.1007/11494744_10 

Abstract: Current modeling techniques are not well equipped to design dynamic software architectures. In this work we present the basic concepts for a dynamic architecture modeling using nets-within-nets. Nets-within-nets represent a powerful formalism that allows active elements, i.e. nets, to be nested in arbitrary and dynamically changeable hierarchies. Applying the concepts from nets-within-nets, therefore, allows us to model complex dynamic system architectures in a simple way, which enables us to design the system at different levels of abstractions using refinements of net models. Additionally to the conceptual modeling of such architecture, we provide a practical example where the concept has been successfully applied in the development of the latest release of Renew (Version 2 of the multi-formalism Petri net IDE). The overall monolithic architecture has been exchanged with a system that is divided into a plug-in management system and plug-in’s that provide functionality for the users. By combining plug-ins the system can be adapted to the users’ needs. Through the introduction of the Petri net concepts, the new architecture is now – at runtime – dynamically extensible by registering plug-ins with the management system. The introduced architecture is applicable for any kind of architecture but most suitable for applications with dynamic structure.

Abstract: In this work we present the basic concepts for a dynamic plug-in-based software architecture using concepts from the Petri net-based MAS framework MULAN. By transferring the concepts of agent-orientation to a plug-in-based architecture we are able to design our application and the plug-in-based system on an abstract level. Moreover, general problems that evolve from a highly dynamic and configurable architecture have been solved by basing the conceptual design on multi-agent principles. The introduced concepts have been applied to the software architecture of the tool RENEW.

Abstract: Destinations are at the heart of travel decisions, and destination image has a significant influence on tourists’ decision-making. Many travelers acquire information via the Inter-net, which offers abundant information and an increasing number of tourism-related services. The impact of media coverage on destination image has attracted research attention and became particularly evident during the 2003 outbreak of SARS, the Severe Acute Respiratory Syndrome. Building upon previous research, this paper analyzes the prevalence of tourism destinations among 162 international media sites. Measuring term frequency investigates the attention a destination receives – from a general and, after contextual filtering, from a tourism perspective. Calculating the semantic orientation estimates positive or negative media influences on destination image at a given point in time. By detecting associations with country names, keyword analysis reveals the countries' public profile, and the impact of events on media coverage.

DOI: https://doi.org/10.1007/11494744_18 

Abstract: The concept of mobile agents imposes a great security risk for information systems. In this paper we propose object nets as a specification formalism for multi-agent systems. Since the general formalism is Turing-powerful not every analysis method that is common for Petri net can be applied. So, we define the subclass of “ordinary” object nets that allows for the application of standard P/T-net techniques, i.e. the computation of boundedness, liveness etc.

Abstract: The e-learning platform K3 realizes a constructivist learning model augmented with collaborative properties. K3 courses, mainly offered since 2004 at the University of Konstanz, follow the blended learning model. K3 collaborative discourse work is organized in virtual groups. All group members have to choose a role (moderator, summarizer, etc.) for a certain period and their role performance is part of their evaluation. Discourse takes place in an electronic (asynchronous) forum. Each contribution/comment must be specified according to its discourse function. These specifications structure discourse and allow selective retrieval of discourse objects. Students are encouraged to augment their contributions informationally by reference objects. A graphic interface facilitates navigation through complex discourse structures and makes them transparent. The technical basis of K3 is an open source, object-oriented client-server system for the management of the different types of K3 data.

DOI: https://doi.org/10.1007/11550648_16 

Abstract: Nowadays many software development (SD) projects are placed in a distributed setting, concerning both the software itself and the resources, processes and actors needed to create or maintain the software. Therefore, tools and methods to support software engineering should be distributed as well. In a SD project many different actors play different roles, all interacting with one another. The software engineering paradigm most suited for this kind of organization in which autonomous actors act and interact is the paradigm of agent-oriented software development (AOSE). This article presents a MAS-based tool platform (cooperative infrastructure) which integrates different tools that are distributed over several agent platforms. The eventual goal of this platform is to create a distributed software development environment. This should easily allow interaction, coordination and cooperation between different participating parties in a software development process, by allowing them to communicate, negotiate, synchronize resources, etc. in a transparent, distributed and dynamic setting.

Abstract: While software archives have been around for a long time, they have been mostly used to store and reconstruct versions of a software system and to coordinate simultaneous changes. Currently many researchers are investigating new ways to exploit the information stored in software archives. In this paper we present an Eclipse plugin that visualizes evolutionary coupling between files, i.e., how likely it is that two files are changed together. The information is automatically extracted from the software archive and displayed as a pixelmap, where each pixel represents a single coupling. The plugln nicely integrates with the rest of the Eclipse IDE and thus allows to interactively explore the evolutionary coupling between different files of a project.

Abstract: Angesichts der gegenwärtigen Umbrüche im Suchdienstemarkt versucht dieser Beitrag Entwicklungstendenzen bei Websuchdiensten aufzuzeigen. Ausgehend von einer Darstellung grundlegender Problemfelder werden neue technologische Ansätze in den Themenfeldern Erschließung neuer Wissensbestände und Ausgabe von Suchergebnissen angeführt. Es wird deutlich, dass die Suchdienste zunehmend Inhalte des Deep Web zugänglich machen. Die großen, international bedeutenden Suchdienste integrieren hierzu zunehmend spezielle Wissensbestände und Suchoptionen, insbesondere Produktsuchdienste und geografische Suchdienste. Damit entwickeln sich diese Dienste in einer steigenden Anzahl von Fällen zu den geeignetsten Anlaufstellen zur Befriedigung von Informationsbedürfnissen. Damit diese technologischen Fortschritte, angesichts der geringen Nutzungskompetenz der Anwender, wirksam werden gehen die Suchdienste dazu über, neue Ansätze bei der Ergebnisausgabe umzusetzen, um dem Nutzer die Auswahl der bestgeeigneten Ergebnistypen aus den verschiedenen vorhandenen Indizes abzunehmen. Voraussetzung dafür ist, dass die Suchdienste in der Lage sind die jeweiligen Informationsbedürfnisse der Nutzer zu erkennen. Personalisierung gilt momentan als Schlüsseltechnologie, um dieses Ziel zu erreichen. Somit deutet sich an, dass sich der Fokus der technologischen Entwicklung langsam wandelt, weg von einer Fixierung auf optimale Rankingverfahren, hin zu einer grundlegenderen Orientierung an den subjektiven Bedürfnissen des Nutzers.