Editorial

The Computer Science and Information Systems journal, in its first issue of Volume 14, consists of nine regular articles and the Special Section on “Ambient Intelligence for Large Premises”, guest-edited by Marin Lujak and Rubén Fuentes. We express our gratitude to the guest editors, authors and reviewers, whose dedication and hard work made the current issue of our journal possible.

The issue’s regular article section begins with “Smart Tourist Information Points by Combining Agents, Semantics and AI Techniques,” authored by Piedad Garrido et al. The paper presents an integrated solution for smart tourist information points that uses a lightweight OWL ontology to describe relevant information on a region, combining it with the Maxine platform to obtain meaningful and informative natural-language interaction with the user by way of artificial intelligence techniques and models.

“ResXplorer: Revealing Relations between Resources for Researchers in the Web of Data” by Laurens De Vocht et al. presents ResXplorer, a search interface that interactively visualizes linked data of research-related sources. ResXplorer is shown to successfully reveal relationships between researchers and resources (conferences, publications, proceedings) that may not be evident from classic faceted search, keyword matching and filtering features offered by current research collaboration platforms such as Mendeley and ResearchGate.

José A. Ruipérez-Valiente et al., in “Evaluation of a Learning Analytics Application for Open edX Platform,” propose a learning analytics application ANALYSE for massive open online courses (MOOCs), and evaluate its usability and effectiveness on the Open edX platform, with students enrolled in the Design of Telematics Applications course, demonstrating good system usability in terms of the usefulness of visualizations and the effectiveness of the actions required for the respondents.

In their article entitled “Investigation of Moderator Factors in e-Business Adoption: A Quantitative Meta-analysis of Moderating Effects on the Drivers of Intention and Behavior,” Boštjan Šumak et al. overview existing knowledge in the field of e-business technology acceptance and explore the differences in Technology Acceptance Model (TAM) related causal effect sizes for different e-business contexts. The results of the meta-analysis of existing empirical research show a moderating effect for all four proposed factors: consumer type, device type, continent and respondent type.

Jelena Graovac et al., in “Hierarchical vs. Flat N-gram-based Text Categorization: Can We Do Better?”, explore the hierarchical text categorization (HTC) problem, presenting and evaluating two HTC techniques based on k-nearest neighbor (kNN) and support vector machine (SVM) classifiers, combined with the byte n-gram document representation. The effectiveness of the presented techniques and their language independence are experimentally demonstrated on five tree-structured benchmark data sets, exhibiting state-of-the-art performance.

In “Generalized Ensemble Model for Document Ranking in Information Retrieval,” Yanshan Wang et al. propose a generalized ensemble model (gEnM) for document ranking. Multiple document retrieval models are combined through an optimization problem formulated to directly maximize the mean average precision, solving it by means of several supervised and unsupervised approaches, making gEnM a viable option in diverse practical information retrieval applications.

The paper “Fast Edge-preserving Gravity-like Image Interpolation” by Željko Lukač et al. proposes a novel image interpolation algorithm which preserves edges and keeps a natural texture of interpolated images, based on the idea that only pixels that belong to the same side of an edge should be used in interpolation of pixels that belong to an edge. Experimental comparison of the proposed method with existing interpolation algorithms demonstrates a favorable tradeoff between obtained image quality and invested computing resources.

Girūta Kazakevičiūtė–Januškevičienė et al., in their article “Annotation of the Medical Images Using Quick Response Codes,” propose a method for medical image annotation based on QR codes. It is shown that the method compares favorably to JPEG 2000 in terms of speed and capacity.

The final regular article, “Methodology to Investigate BitTorrent Sync Protocol,” by Algimantas Venčkauskas et al. proposes a methodology that enables capturing the initiation step of the BitTorrent Sync client application, and informing the forensics investigator of possible illegal content being shared. Capabilities of the method were demonstrated on simulated use of BitTorrent Sync, as well as on monitored real Internet traffic.

Editor-in-Chief
Mirjana Ivanović

Managing Editor
Miloš Radovanović