Since the announcement of the two-year impact factor of our journal in the summer of last year, the number of submissions to ComSIS has increased dramatically, placing a strain on our editorial staff. In order to continue the publication of our journal in a timely and quality manner, we increased the number of editorial rejections, as well as strengthened our criteria for acceptance of articles for publication. If this trend continues, we may additionally be forced to adopt the policy of accepting only a certain portion of articles (undeserving of editorial rejection) into the reviewing process. We apologize to all prospective authors for any inconvenience this may cause, but the circumstances and our commitment to keeping and enhancing the quality standard of our journal are influencing this shift in policy.

In addition, we are happy to announce that the two-year impact factor of our journal has increased from 0.324 (2010) to 0.625 (2011).

On behalf of the Editorial Board and the ComSIS Consortium, we thank our Guest Editor, Prof. Vladan Devedžić (University of Belgrade, Serbia) for organizing the special section on emerging trends in technology-enhanced learning (TEL), featured in this issue of ComSIS. We would also like to express our gratitude to the authors and the reviewers for their high-quality contributions and effort invested into this issue of our journal.

This issue of Computer Science and Information Systems consists of fifteen regular articles and five articles in the special section on TEL.

In the first regular article, “Building XML-Driven Application Generators with Compiler Construction Tools,” Antonio Sarasa-Cabezuelo, Bryan Temprado-Battad, Daniel Rodríguez Cerezo and José-Luis Sierra describe the use of classic compiler-construction tools (i.e., parser generators), to build XML-driven application generators. Their approach consists of a document interface based on standard stream-oriented XML processing, used to build the XML scanner, characterization of the syntax of the streams by generation-specific context-free grammars, and augmentation of these grammars with suitable semantic attributes and semantic actions.

The next paper by Dragan Milićev, “Towards Understanding of Classes versus Data Types in Conceptual Modeling and UML,” provides an in-depth study of the ambiguities and discrepancies when making the distinction between two kinds of entity types – classes and data types – in traditional conceptual modeling and UML. A novel semantic interpretation is proposed for consolidation, based on the premise that populations of the two kinds of entity types are defined in two ways: by intensional (for data types) and extensional (for classes) definitions. The proposed interpretation also lends itself to the description of several semantic consequences: value-based vs. object-based semantics, associations vs. attributes, and identity vs. identification.

“Specification of Data Schema Mappings using Weaving Models”, by Nenad Aničić, Siniša Nešković, Milica Vučković and Radovan Cvetković, tackles the problem of unsuitability of weaving models for the specification of heterogeneous schema mappings in model driven engineering. The lack of mapping rules, which allows specifications that are semantically meaningless, wrong, and/or disallowed, is overcome by providing the explicit support for semantic mapping rules, based on the introduction of weaving metamodels augmented with constraints.

Vıctor M. Prieto, Manuel Alvarez, Rafael Lopez-Garcıa and Fidel Cacheda, in “A Scale for Crawler Effectiveness on the Client-Side Hidden Web,” present an evaluation framework for assessing the success of crawlers in traversing the client side of the “hidden Web.” The authors formulate a scale by grouping basic scenarios in terms of several common features, and propose evaluation methods of crawler effectiveness in terms of the scale’s levels. Through extensive evaluation of open-source and commercial crawlers, the paper highlights the ones that are most capable of handling the evaluation scenarios.

In “Wavelet trees: a survey,” Christos Makris gives an overview of various characteristics of the wavelet tree data structure, which is used for many tasks, such as text compression, text indexing, and retrieval. The article considers issues concerning the efficient maintenance of the structure, and its handling in various applications.

The article “Impact of Personnel Factors on the Recovery of Delayed Software Projects: A System Dynamics Approach,” by Mostafa Farshchi, Yusmadi Yah Jusoh and Masrah Azrifah Azmi Murad, explores the possibility for significant schedule improvement by adding new staff during delayed realization of a software project, considering new manpower’s capabilities, skills and experience. The study is conducted through formulation and evaluation of a system dynamics model which simulates the project’s progress when new members are added.

“A New Strategic Tool for Internal Audit of the Company Based on Fuzzy Logic,” by Aleksandar Pešić, Duška Pešić and Andreja Tepavčević, presents strategic management tool for the assessment of internal organizational factors. The tool addresses the limitations of traditional appraisal methods, enabling more comprehensive evaluation of a company’s internal environment. The authors propose an original method – fuzzy synthesis of internal factors (FSIF) – which extends the approach of the classical internal factor evaluation (IFE) matrix approach by incorporating fuzzy logic in order to better describe real situation.

In the paper “Experimental Investigation of the Quality and Productivity of Software Factories Based Development,” Andrej Krajnc et al. evaluate the quality and productivity benefits of the software factories (SF) approach to software development. The traditional and the SF approach were compared using the Goal – Question – Metric (GQM) methodology, on participants grouped into 32 teams, demonstrating the superiority of the SF method.

“Nearest Neighbor Voting in High Dimensional Data: Learning from Past Occurrences,” by Nenad Tomašev and Dunja Mladenić, presents a new approach for exploiting the hubness phenomenon, which is an aspect of the “curse of dimensionality” inherent to nearest-neighbor (NN) methods, in k-NN classification. The article proposes a novel algorithm, hubness information k-nearest neighbor (HIKNN), which introduces the informativeness of a point’s hubness into the k-NN voting framework.

In “AKNOBAS: A Knowledge-based Segmentation Recommender System based on Intelligent Data Mining Techniques,” Alejandro Rodríguez-González et al. describe design and evaluation of the recommender system AKNOBAS, which employs clustering algorithms for the tasks of marshalling information and recommending content to users.

The paper “Scope of MPI/OpenMP/CUDA Parallelization of Harmonic Coupled Finite Strip Method Applied on Large Displacement Stability Analysis of Prismatic Shell Structures,” by M. Hajduković et al., studies the effectiveness of parallelization in large displacement stability analysis of orthotropic prismatic shells. The computational overhead introduced by the semi-analytical harmonic coupled finite strip method (HCFSM) used to solve the large deflection and the post-buckling problems is tackled by application of MPI, OpenMP and CUDA approaches to parallelization.

Yordan Kalmukov, in the article “Describing Papers and Reviewers’ Competences by Taxonomy of Keywords,” deals with the problem of automatically assigning reviewers to articles by introducing a taxonomy of keywords to describe the competencies of both the articles and reviewers. The hierarchical nature of the taxonomy allows the expression of semantic proximity in addition to exact matches. A modification of Dice coefficient is proposed to achieve the given task.

“Journal Evaluation based on Bibliometric Indicators and the CERIF Data Model,” by Dragan Ivanović, Dušan Surla and Miloš Racković, proposes an application of extended CERIF data model for storing journal impact factors and journal scientific fields, and based on this data presents a journal evaluation approach. The method does not unambiguously evaluate a journal, but rather suggests possibly journal appropriate categories according to the values of the employed metric, but final decision is made by a committee.

In the paper “Ontology-Based Home Service Model,” Moji Wei, Jianliang Xu, Hongyan Yun and Linlin Xu study home service retrieval and invocation for a smart home. An ontology-based service model is employed to retrieve and invoke services according to user’s needs automatically. Two domain ontologies are constructed – function concept ontology and context concept ontology – to annotate the semantics of the home service different facets. Two scenarios involving different types of services are presented that demonstrate the usage of the model.

Finally, in “Automatic Generation of E-Courses Based on Explicit Representation of Instructional Design,” Goran Savić, Milan Segedinac and Zora Konjović present a system for automatic generation of IMS LD compliant e-courses from three components: machine readable explicit representation of instructional design, ontology of learning goals, and IMS content packaging compliant learning resources.

Mirjana Ivanović

Managing Editor
Miloš Radovanović