Editorial

Beginning Volume 13 of the Computer Science and Information Systems journal, this issue consists of 14 regular articles, and (for the first time) two errata. We acknowledge the hard work and enthusiasm of all authors and reviewers, whose efforts were paramount for the creation of our journal‟s current issue.

This issue starts with the article by Nenad Tomašev, Krisztian Buza and Dunja Mladenić, “Correcting the Hub Occurrence Prediction Bias in Many Dimensions”, where the bias of instance selection from a data set is examined through the lens of hubness – the property of k-nearest neighbor graphs constructed from high-dimensional data using a distance measure to contain highly influential and connected nodes. It is revealed that different instance selection strategies bias the predictions of the behavior of hub-points in different ways. The paper proposes an un-biasing step to improve various hubness-aware classification methods.

Petar Vračar, Igor Kononenko and Marko Robnik-Šikonja, in “Obtaining Structural Descriptions of Building Façades” propose a method for recognizing windows as basic structural elements of building façades, and constructing an interpretable façade model. A likelihood map of window locations is created, and explained by a structural façade model based on a formal grammar. The model is obtained by searching the grammar space using a look-ahead greedy search method.

“S-TFRC: An Efficient Rate Control Scheme for Multimedia Handovers” by Imdad Ullah, Zawar Shah and Adeel Baig tackles the challenges of integrating heterogeneous wireless access technologies, where handover can significantly impair the performance of multimedia applications. The article proposes a state-aware feedback extension to the datagram congestion control protocol (DCCP) that meets QoS requirements of multimedia applications throughout the handover process. Movement of mobile subscribers from unstable to more stable environments is modeled using different distributions, and a Markov model is constructed capturing TCP timeouts, loss rates, and no-feedback timer expirations, with the ability to predicts various congestion events during handover.

In their article entitled “Packet Dispersion Strategy Evaluation from the Perspective of Packet Loss Pattern and VoIP Quality,” Nemanja Ninković, Slavko Gajin and Irini Reljin address the problem of burst packet losses and their devastating effect on Voice over IP (VoIP). The authors introduce strategies to use packet dispersion for loss recovery, providing a computational model, and testing it in a real network environment. It is demonstrated that dispersion is qualitatively superior to single path routing, and that different dispersion strategies perform differently under various burst loss settings.

Xian Guo et al., in “Publish/Subscribe Scheme For Content Centric MANETs”, provide a solution to the problem introduced when data and data source are decoupled, allowing caching of data anywhere in a mobile ad hoc network (MANET). Such decoupling makes ensuring that content publishing or subscribing is done only by legitimate users a challenge. The authors propose a content publish/subscribe system based on distributing “licenses” to legitimate users. The scheme also helps in defending networks against common attacks, e.g. interest flooding and content pollution.

The paper “Preserving Source- and Sink-location Privacy in Sensor Networks” by Sangho Lee, Jong Kim and Yoonho Kim tackles the problem of protecting the location privacy of source and sink nodes in a sensor network. The authors propose a constant-rate broadcast scheme for ensuring location privacy of sources and sinks, which equalizes traffic patterns of the sensor network to deal with eavesdropping and minimizes the routing information of each sensor node to deal with node compromising.

Sanghyun Park et al., in their article “Efficient Processing of Continuous Queries Utilizing F-relationship in Stock Databases,” optimize the process of query computation in a stock investment recommender system by determining interdependencies between query decisions, and using them to maximize the number of queries that do not need to be computed. This is achieved by defining a relation between queries, F-relationship. An F-relationship between queries Q1 and Q2 means that, if the recommendation type of a preceding query Q1 is X, then its following query Q2 always has X as its recommendation type, where X is one of SELL, HOLD, BUY, and NONE. Experiments demonstrate massive improvement of overall time of continuous query processing with the proposed methods compared to the traditional baseline.

In “Current Prospects Towards Energy-Efficient Top HPC Systems,” Sonja Filiposka, Anastas Mishev and Carlos Juiz review the state-of-the art in high-performance computing (HPC) from the perspectives of both performance and power consumption. Using historical information from the Top and Green HPC lists, the article points to the most promising design options and their combinations, revealing the significant challenges for future research.

The next article, “A Systematic Data Collection Procedure for Software Defect Prediction” by Goran Mauša, Tihana Galinac Grbac and Bojana Dalbelo Bašić tackles the challenges raised by using different data repositories in software defect prediction research. The article identifies some of the issues regarding the minimization of dataset bias, and provides measures that can help researchers compare data collection approaches and evaluate data quality. Furthermore, the article proposes a data collection procedure that uses a bug-code linking technique based on regular expressions.

Valentino Vranić and Milan Laslop, in “Aspects and Roles in Software Modeling: A Composition Based Comparison,” discuss the similarities between aspect-oriented and role-based composition and decomposition in software modeling. Using Theme/UML (a well-known approach to aspect-oriented modeling) and OOram (prominent approach to role-based modeling that influenced UML), the authors establish a partially reversible transformation of a Theme/UML model to the corresponding OOram model that strongly supports the observed analogy. Based on this, the article sketches the possible use of UML for aspect-oriented modeling.

“Computer Aided Anonymization and Redaction of Judicial Documents,” by Goran Sladić et al. addresses the reconciliation of the need for public access to case law, and judgments in particular, with preservation of privacy and personal integrity. The paper proposes a computer-aided method for anonymization and redaction of judgments, based on the access control mechanism for XML documents, with the aim to improve the efficiency of the overall process.

In “A Methodology for Determining the Value Generation Mechanism and the Improvement Priorities of Open Government Data Systems,” Charalampos Alexopoulos, Euripides Loukis and Yannis Charalabidis analyze what value open government data (OGD) systems create, and identify the main improvements these systems require. The methodology proposed for this task is based on the estimation of a „value model‟ of the OGD system under evaluation from users‟ ratings, which consists of several value dimensions and their corresponding value measures, organized in three „value layers‟, and also the relations among them.

Qifang Luo, Mingzhi Ma and Yongquan Zhou, in their paper “A Novel Animal Migration Algorithm for Global Numerical Optimization,” improve the animal migration optimization (AMO) algorithm for numerical optimization by proposing a novel migration process to improve the exploration and exploitation ability of the animal migration optimization. The proposed algorithm compares favorably to the state-of-the-art methods, including particle swarm optimization (PSO) and cuckoo search (CS).

The final article of this issue by Zhang Tingting et al., “Capability-oriented Architectural Analysis Method Based on Fuzzy Description Logic,” presents a method for modeling architectural requirements through ontology-based and capability-oriented requirements elicitation, with fuzzy UML introduced to model the vague and uncertain features of the capability requirements.

For the first time in Computer Science and Information Systems, this issue includes two errata, with corrections to author affiliations in previously published articles: “Exploiting Geotagged Resources to Spatial Ranking by Extending HITS Algorithm” by Tuong Tri Nguyen and Jason J. Jung (Vol. 12, No. 1, pp. 185–201, 2015), and “The Use of Ontologies in Cadastral Systems” by Dubravka Sladić et al. (Vol. 12, No. 3, pp. 1033–1053, 2015).

Editor-in-Chief
Mirjana Ivanović

Managing Editor
Miloš Radovanović