Search

Additional considerations regarding (a) how the data model presented in “Results” section recognizes the design principles of extensibility and data lineage introduced in “Methods” section; and (b) variant identification for sematic integration when aggregating data from multiple source databases. (PDF 95 kb)

The social web has become a major repository of social and behavioral data that is of exceptional interest to the social science and humanities research community. Computer science has only recently developed various technologies and techniques that allow for harvesting, organizing and analyzing such data and provide knowledge and insights into the structure and behavior or people on-line. Some of these techniques include social web mining, conceptual and social network analysis and modeling, tag clouds, topic maps, folksonomies, complex network visualizations, modeling of processes on networks, agent based models of social network emergence, speech recognition, computer vision, natural language processing, opinion mining and sentiment analysis, recommender systems, user profiling and semantic wikis. All of these techniques are briefly introduced, example studies are given and ideas as well as possible directions in the field of political attitudes and mentalities are given. In the end challenges for future studies are discussed.

An overview of the Large-Scale Multi-Agent Modelling of Massively Multi-Player On-Line Role- Playing Games (ModelMMORPG) project financed by the Croatian Science Foundation is provided. ModelMMORPG aimed on studying Massively Multiplayer Online Role- Playing Games (MMORPGs) using a combined computer science Multiagent System (MAS) and a social science Agent-Based Modelling (ABM) methodology. The main results, performed activities, developed software as well as published papers are described and discussed. In the end a critical evaluation of the project is given.

In the last decade advances of computer technologies have lead towards a technological reality where the line between information consumers and information producers is blurred. This technological omnipresence allows for unprecedented data creation capabilities. Based on various data sources, it seems humans have fully embraced data- generating activities. One such activity is using online social network applications, like Facebook or Twitter in almost all aspects of their lives. One of the main features of online social network applications is perceived freedom of speech, individuality and privacy, even though every application has some special features. Therefore, content generated using these services presents the public with interesting insights in private life of people and their attitudes towards public affairs. Social network applications are active the most during specific public events aimed at the massive public. Due to its brevity, ease of use and frequency, Twitter is an interesting social network application for research and analysis. Other than allowing almost exclusively short messages (up to 140 characters), a tweet (a Twitter post) can contain location of the message sender as well as a graphic to accompany the textual message. The textual part of the message may contain so called hashtags – keywords used for indexing and easy identification of a subject the message is related to. These hashtags allow us to group messages related to a specific event. Recent governmental elections held in Croatia were very popular amongst the Croatian Twitter community. Usage of hashtags allowed us to identify the right messages and thus most- used words to describe this event and potentially identify how people felt when talking, i.e. writing, about politics and the held elections. Furthermore, geolocation information, optionally embedded in a tweet, makes it possible to analyze which keywords were used in which parts of Croatia, all pertaining to the held governmental elections. Public opinion is only a few tweets away, but are the results similar to the election results? If tweet-based opinion can be constructed, does it differ from the real results? These and similar questions will be addressed in this paper.

The thesis establishes a new technique for web news portal content personalization based on (1) extracting individual behavioral patterns of users, (2) information retrieval techniques and (3) weighted Voronoi digrams. The aim of this work is to test the existing methods, which are identified to be of importance for future steps, and to improve these methods with the help of weighted Voronoi diagrams. The purpose of using Voronoi diagrams in this process is to achieve personalization at the individual level, differing from the standard clustering approaches. Using weighted Voronoi diagrams for personalization has so far not been reported in the literature and therefore represents a novel approach to the process of information personalization. Weighted Voronoi diagrams are suitable for this purpose as they allow for standardized distance metrics expanded with personalization parameters to be used.

The main objective of this study is to obtain a holistic view of Cloud Computing ontologies, their applications and focuses. The identification of primary studies in this systematic review is based on a pre-defined research protocol with a research question, inclusion and exclusion criteria and a search strategy. We summarize the selected studies into four main categories: Cloud resources and services description, Cloud security, Cloud interoperability and Cloud services discovery and selection. The analysis of the included studies indicates a number of challenges and topics for future research, including those specifically related to using ontologies to improve security and interoperability of Cloud Computing offerings.

Age estimation of humans is one of the problems in field of computer vision which is insufficient researched. To efficiently estimate the age of an individual based on his/her face image, characteristic points of the face have to be determined. In this paper the most common face recognition algorithms will be described.

The definition of information crisis notices the problem of the information age: too much information which in return makes it hard to get to the needed information when needed. Almost every Internet user has experienced this effect of redundant amount of information delivered to him that can cause missing out on the needed information as well as repetition of already digested information. The gathering of information also includes data aggregation from more than one source which causes repetition of already seen information. In this paper we present a system that uses a user friendly method for the creation of website templates which define a HTML wrapper. Such system serves as information backbone for other developers which want to use data from other websites that do not offer any APIs. The possibilities of such a system are limitless for today’s Semantic Web applications as well extended media Services.

Semantic wiki systems have gained major attention of the academic community in the past few years. The idea of integrating the semantic web with fast and lightweight content management systems seems to be good direction towards web 3.0 or the social semantic web. Herein we analyze how to take advantage of such systems through dynamic querying facilities. Particularly we analyze the \taopis system which has a build in frame logic reasoning engine, and show how queries can enhance the presented content through dynamically generated class hierarchies, dictionaries, FAQ-s, tables of content, issue tracking and others. We conclude that by enriching wiki syntax with dynamic queries semantic wikis can gather additional functionality in organizing knowledge.

In this paper we present the TaOPis semantic wiki system based on F-logic that we implemented. The system uses the Flora-2 reasoning engine and supports the dynamic creation of queries in restricted Flora-2 syntax. We show how we were able to apply a simple tagging system in order to support the creation of an object oriented ontology. In the end we give a comparison between other semantic wiki systems and the developed system as well as give guidelines for future research.

With the development of the Web and the services it provides to the end user one of the aspects one has to take in mind is the need for an account for each of its services. Until now the accounts used a username and password model for the user authentication but with the development of the biometric science and high-speed Internet there are other possibilities that provide a more secure solution. Instead of using username and password we can use behavioral or psychophysical biometrical characteristics to prove if the person is a legal or illegal user. Many today's computers or notebooks have fingerprint readers (usually on mouse) or have some cameras as an addition to standard IO (input/output) devices, keyboard and mouse. Using only these two characteristics a model for so called e-Authentication, based on multimodal biometrics systems, can be developed. This multimodal biometrics system can be supported by soft biometrical characteristics (like eye color or hair color) or some other biometrics system like keystroke dynamics. Possible applications include securing password managers, points of single logon etc. In this paper we described the development of multimodal biometrics system, called eBAWS (e- Biometric Authentication for Web Systems), based on standard IO devices based on keystroke dynamics, shape drawing, shape order and handwriting verification.