I am honored and privileged to take the Editor-in-Chief position in ComSIS for Vol.2, issues 1 and 2. I am most grateful to Professor Lazarević, the previous Editor-in-Chief, for putting his trust in me and nominating me for that position. It was a great feeling to receive encouragement from all the other members of the ComSIS consortium, who supported my nomination and provided the necessary advice and cooperation.

It is not only me – from this issue, the new Vice Editor-in-Chief is Dr. Ivan Luković from the University of Novi Sad, and the new Editorial Assistant is Sonja Radenković, a young graduate student from the University of Belgrade. Dr. Jelica Protić, also from the University of Belgrade, is a new member of the ComSIS Editorial Council. I welcome them on board and strongly believe that the experience they will have with ComSIS will be as great as mine.

I am also extremely happy to note that ComSIS seems to have got recognition wider than we originally expected. For example, we have recently received a request from UNESCO Institute for Information Technologies in Education (IITE) to grant permission to them to use excerpts from a high-quality ComSIS paper (by E. Duval) in their global Cross-cutting Theme Project Methodologies for Digital Libraries (CCTP MDL). Also, there is a constant inflow of submissions to ComSIS from different countries.

The third issue of ComSIS is much like the first two in terms of organization and structure. As in the first two issues, here we also have two invited papers and several regular research papers. The two invited papers come from two distinguished scientists in the field of computing, Professor John Debenham from the University of Technology, Sydney, Australia, and Professor Riichiro Mizoguchi, from the Osaka University, Japan.

The paper by J. Debenham opens the issue by presenting a strong formal model of knowledge base maintenance. It starts from the concept of an item, an abstract unit of knowledge representation in a knowledge base. In order to maintain a knowledge base, it is important to understand what has to be done to maintain the integrity of its items. Maintaining the integrity of the items in a knowledge base leads, in turn, to a specification of the modifications to the set of programs that implement it. As the paper shows, it turns out that the maintenance of the formal model can be analyzed by introducing the concept of maintenance links. Professor Debenham's long-time work on formal representation of knowledge has earned him the reputation of a key figure in artificial intelligence today.

The second invited paper in this issue, by R. Mizoguchi, comes in the form of an interview. Dr. Jacqueline Bourdeaux from Télé-université, Montréal, Canada, did a great job of interviewing Professor Mizoguchi on the topics of ontologies and ontological engineering. The outcome is more than an exciting (albeit personal) view of the field of ontologies, which lays the foundation of and the infrastructure for the Semantic Web. As the paper states explicitly, different people have different understanding of ontologies and their importance in computing, so the paper clarifies a number of interesting issues to this end. Professor Mizoguchi's contributions to the field of ontologies are widely known and make him one of the world's leading scientists in that discipline.

Lj. Kašćelan and D. Bečejski-Vujaklija tackle in their paper the important problem of financial crisis management. They present a brief analysis of the financial crisis management process and identify key business processes involved. Then they describe a model of financial crises symptoms, a data mining algorithm for automatic detection of such symptoms, and a related practical implementation.

The paper by A.N. El-Kassar and R.A. Haraty is another paper in this issue with a strong formal approach. The paper's topic areas are cryptography and cryptosystems. The authors propose an extension of the well-known ElGamal encryption scheme, and analyze the advantages of such an extension.

V.Vranić 's paper is about software design and programming language paradigms. Each specific paradigm has its advantages and disadvantages, relative to the problem being solved, hence the ideas of combining several paradigms into one – multiparadigm design and multiparadigm programming language. V. Vranić proposes a method called multiparadigm design with feature modeling for selecting a paradigm appropriate for the problem being solved, using the aspect-oriented approach.

Z. He et al. present a new method for the data mining task called outlier detection. Outliers are defined as the data transactions that contain less frequent patterns in their itemsets, and their method detects outliers by discovering frequent patterns (or frequent itemsets) in data sets. The paper also defines a measure called FPOF (Frequent Pattern Outlier Factor) to detect the outlier transactions and describes the FindFPOF algorithm for outlier discovery.

N Aničić and N. Ivezić analyze Semantic Web technologies for enterprise application integration standards. They realize that large-scale, industry-wide enterprise integration efforts depends on the enterprise application integration standards, but also that current XML-syntax-based standards provide only limited reasoning capabilities. Hence they evaluate capabilities of the Semantic Web technologies for enterprise application integration and, in particular, how such technologies affect integration testing.

Last but not the least, I would like to announce the introduction of the practice of occasional publication of ComSIS special issues, dedicated to specific areas of computing. Vol.2 issue 4 will be the first such an issue, and will focus on e-Learning.

Vladan Devedžić