Search

publishedVersion

Digitalization is introducing a virtual representation of the physical world. Such virtual or digital assets, interconnected via global communication network, open unbounded management opportunities that are realized in form of software. Nowadays, digitalization is not a matter of technology change, but is rather perceived as a prerequisite of modern and sustainable operation. Thus, proactive digitalization continuously increases the complexity of globally interconnected software systems and new emerging technologies may bring significant improvements in their management. Management and sustainable goals depend on sound and effective software system engineering solutions. In this paper, we discuss modeling complex software systems within a water supply case study, in which digitalization is introduced to accomplish sustainable goals. Our focus is on water leak report and repair process, for which we open interesting further research and development challenges, based on our previous experiences in modeling complex software systems and in modeling software failure report and repair process.

Software build and integration procedures are responsible to transform developers’ files into machine executable codes. This process separates the software development project from the software deployment phase and it represents a critical obstacle for software performance on the software developer side and deployment effectiveness from the software execution side. Although the automation of this process is an active research topic for almost half a century, the standardized procedures that would enable full automation across various software projects, but also across time, are still missing. The main reason lies in the variety of methods and tools used in software development and their evolution. Therefore, many researchers and practitioners report a low build success rate. Researchers have investigated this problem from software development, integrator, and software evolution perspectives. In this paper, a study is set up in order to contribute to the overall understanding of the level of standardization among FOSS software repositories and automation abilities. A rich source of information lies in FOSS repositories and mining these repositories with joined developers and deployment perspective would bring valuable knowledge into the software engineering profession.

This thesis presents an approach for working with low level source code that enables automatic restructuring and raising the abstraction level of the programs. This makes it easier to understand the logic of the program, which in turn reduces the development time.The process in this thesis was designed to be flexible and consists of several independent tools. This makes the process easy to adapt as needed, while at the same time the developed tools can be used for other processes. There are usually two basic steps. First is the translation to WSL language, which has a great number of semantic preserving program transformations. The second step are the transformations of the translated WSL. Two tools were developed for translation: one that works with a subset of x86 assembly, and another that works with MicroJava bytecode. The result of the translation is a low level program in WSL.The primary goal of this thesis was to fully automate the selection of the transformations. This enables users with no domain knowledge to efficiently use this process as needed. At the same time, the flexibility of the process enables experienced users to adapt it as needed or integrate it into other processes. The automation was achieved with a hill climbing algorithm.Experiments that were run on several types of input programs showed that the results can be excellent. The fitness function used was a built-in metric that gives the “weight” of structures in a program. On input samples that had original high level source codes, the end result metrics of the translated and transformed programs were comparable. On some samples the result was even better than the originals, on some others they were somewhat more complex. When comparing with low level original source code, the end results was always significantly improved., U okviru ove teze se predstavlja pristup radu sa programima niskog nivoa koji omogućava automatsko restrukturiranje i podizanje na više nivoe. Samim tim postaje mnogo lakše razumeti logiku programa što smanjuje vreme razvoja.Proces je dizajniran tako da bude fleksibilan i sastoji se od više nezavisnih alata. Samim tim je lako menjati proces po potrebi, ali i upotrebiti razvijene alate u drugim procesima. Tipično se mogu razlikovati dva glavna koraka. Prvi je prevođenje u jezik WSL,za koji postoji veliki broj transformacija programa koje očuvavaju semantiku. Drugi su transformacije u samom WSL-u. Za potrebe prevođenja su razvijena dva alata, jedan koji radi sa podskupom x86 asemblera i drugi koji radi sa MikroJava bajtkôdom. Rezultat prevođenja je program niskog nivoa u WSL jeziku.Primarni cilj ovog istraživanja je bila potpuna automatizacija odabira transformacija, tako da i korisnici bez iskustva u radu sa sistemom mogu efikasno da primene ovaj proces za svoje potrebe. Sa druge strane zbog fleksibilnosti procesa, iskusni korisnici mogu lakoda ga prošire ili da ga integrišu u neki drugi već postojeći proces.Automatizacija je postignuta pretraživanjem usponom (eng. hill climbing).Eksperimenti vršeni na nekoliko tipova ulaznih programa niskog nivoa su pokazali da rezultati mogu biti izuzetni. Za funkciju pogodnosti je korišćena ugrađena metrika koja daje “težinu” struktura u programu. Kod ulaza za koje je originalni izvorni kôd bio dostupan, krajnje metrike najboljih varijanti prevedenih i transformisanih programa su bile na sličnom nivou. Neki primeri su bolji od originala, dok su drugi bili nešto kompleksniji. Rezultati su uvek pokazivali značajna unapređenja u odnosu na originalni kôd niskog nivoa.

Massively multiplayer online games (MMOGs) and social networks are very popular communication and entertainment formats, where millions of players from all around the world interact in a shared environment and exchange and trade different types of digital media: texts, videos, sounds and music. The communication systems implemented in these virtual environments are in increasing number encrypted to prevent fraud and user impersonation. Ubiquity of virtual forums with massive participation and participants communication systems has raised several questions about their security - gaming servers are suitable as potential exploitation tools for terrorist groups use to conduct on-line operations. The prevailing on Internet top-down security organization where trustworthiness of an object (computer, program) is established by a strong hierarchical system of security certificates does not work, since trusted high-level certificates can already be bought online in dark web marketplaces. For many local communities (players of an online multiplayer game, local social networks etc.) is more advantageous a different, local 'sand-box' organization of a communication system where anonymous participants (initially identified only by their generated username without using even e-mails) actively interact using messages with strong encryption. For encryption they need entropy/randomness, but this they can create themselves in their interactions. In competitive interactions (e.g. they play a competitive game) all participants (trying to compete each other) behave differently, try to create thus the sequence of their actions is random and can be used as the secure key for symmetrical encryption for communication among participants. Here is considered a class of games where expectation of payoff is the same for all moves, thus players cannot get from results any additional information about the game, thus their best strategy is to select their moves randomly (non-learnable games). It is shown, that in a sense all games of this class are similar, can be created with the same procedure and can be reduced to each other using introduced here operation of rectangular modification of the game state matrix. The best strategy for moves in this class of games is uniform randomness, thus (if they are competing and try to beat each other) in the play they create with their moves a random sequence. This sequence can be used for generating a key for symmetric encryption. In a non-local (server-based) multiplayer game players know only their own moves, about moves of other players they get only results of game (not the actual moves made by other players). Thus for generating a key server sends players sequence of all moves from which the player's own moves are removed. This sequence is different for players and contains only partial information, thus an eavesdropper (man-in-the-middle) cannot use it. Players insert into this holey sequence their own moves and get all the same sequence of moves which will be used as the common key for symmetrical encryption of communication; the procedure allows several enhancements for further randomness and/or speed of key generation. The key generation from player's moves removes need for use of public-key systems and all communication (and keys) remain inside the virtual community, whose security thus becomes self-sustainable. publishedVersion

An analysis of fault distribution in software systems was conducted in the classical work by Fenton and Ohlsson. Their study was replicated twice in order to support the findings. They investigated the Pareto principle, fault persistence, relation between code metrics and fault number, as well as fault density and fault density similarities across project releases. In this study, we aim to replicate the study on the data set from completely different environment in order to see if the evidence obtained from the previous studies can be obtained for software systems that were developed in the open source environment. Our study is conducted on two Eclipse projects in evolution involving 12 releases each. The results are in agreement with the previous studies, except that there is a difference in the behavior of fault densities between pre and post release faults.

Дисертација пружа емпиријски утемељен иновативан и јединствен увид у везу играња дигиталних игара са профилом вишеструких интелигенција ученика виших разреда основе школе. Систематизован преглед литературе реализован у оквиру дисертације указује на чињеницу да се истраживачи слажу у ставу да су дигиталне игре постале методички валидан образовни алат, али да постоји веома мали број емпиријских истраживања посвећених вези између играња дигиталних игара и профила вишеструких интелигенција ученика. Истраживање је реализовано на узорку од 1262 ученика узраста од 11 до 15 година из 22 школе у Србији. За потребе истраживања креирани су и евалуирани одговарајући психометријски инструменти. Анализирана је веза профила вишеструких интелигенција ученика виших разреда основне школе са склоностима и навикама играња дигиталних игара. Проучавана је могућност предикције префериране врсте дигиталних игара у зависности од профила вишеструких интелигенција и навика играња дигиталних игара. Утврђено је да је профил вишеструких интелигенција валидан предиктор склоности, навика играња и идентификатор зависности од дигиталних игара. Извршена је идентификација зависности од дигиталних игара и анализирана је њена повезаност са профилима вишеструких интелигенција. Анализирани су утицаји појединих социо- демографских фактора на склоности и навике играња дигиталних игара и зависност од дигиталних игара. У складу са циљем и задацима истраживања формиране су хипотезе, које су дискутоване и потврђене. На основу резултата утврђено је постојање везе играња дигиталних игара са профилом вишеструких интелигенција ученика виших разреда основе школе. Disertacija pruža empirijski utemeljen inovativan i jedinstven uvid u vezu igranja digitalnih igara sa profilom višestrukih inteligencija učenika viših razreda osnove škole. Sistematizovan pregled literature realizovan u okviru disertacije ukazuje na činjenicu da se istraživači slažu u stavu da su digitalne igre postale metodički validan obrazovni alat, ali da postoji veoma mali broj empirijskih istraživanja posvećenih vezi između igranja digitalnih igara i profila višestrukih inteligencija učenika. Istraživanje je realizovano na uzorku od 1262 učenika uzrasta od 11 do 15 godina iz 22 škole u Srbiji. Za potrebe istraživanja kreirani su i evaluirani odgovarajući psihometrijski instrumenti. Analizirana je veza profila višestrukih inteligencija učenika viših razreda osnovne škole sa sklonostima i navikama igranja digitalnih igara. Proučavana je mogućnost predikcije preferirane vrste digitalnih igara u zavisnosti od profila višestrukih inteligencija i navika igranja digitalnih igara. Utvrđeno je da je profil višestrukih inteligencija validan prediktor sklonosti, navika igranja i identifikator zavisnosti od digitalnih igara. Izvršena je identifikacija zavisnosti od digitalnih igara i analizirana je njena povezanost sa profilima višestrukih inteligencija. Analizirani su uticaji pojedinih socio- demografskih faktora na sklonosti i navike igranja digitalnih igara i zavisnost od digitalnih igara. U skladu sa ciljem i zadacima istraživanja formirane su hipoteze, koje su diskutovane i potvrđene. Na osnovu rezultata utvrđeno je postojanje veze igranja digitalnih igara sa profilom višestrukih inteligencija učenika viših razreda osnove škole. Dissertation provides an empirically grounded innovative and unique insight into relations between digital gameplay and multiple intelligences profile. A systematic literature review that was realized in this dissertation points to thefact that researchers agree that digital games became methodically valid educational tool, but that there are very few empirical studies on relations between digital gameplay and multiple intelligences profile of the students. The research sample consisted of 1262 11-to-15- year-old students from 22 Serbian schools. The adequate psychometric instruments were developed and evaluated as part of the research. The relations between multiple intelligence profile of higher grade elementary school students and their gameplay preferences and habits were analyzed. The possibility of predicting preferred digital games genre based on multiple intelligences profile and digital gameplay was examined. It was found that the multiple intelligences profile is a valid predictor of digital gameplay preferences, habits and addiction. The digital games addiction was identified, and its relation with the structure of multiple intelligence profile was examined. The influence of socio-demographic factors on digital gameplay preferences, habits and addiction was analyzed. In accordance with the dissertation objectives and tasks the research hypotheses were formulated, and further discussed and confirmed. The results imply the existence of the relation between digital gameplay and multiple intelligences profile of higher grade elementary school students.

Дисертација пружа емпиријски утемељен иновативан и јединствен увид у везу играња дигиталних игара са профилом вишеструких интелигенција ученика виших разреда основе школе. Систематизован преглед литературе реализован у оквиру дисертације указује на чињеницу да се истраживачи слажу у ставу да су дигиталне игре постале методички валидан образовни алат, али да постоји веома мали број емпиријских истраживања посвећених вези између играња дигиталних игара и профила вишеструких интелигенција ученика. Истраживање је реализовано на узорку од 1262 ученика узраста од 11 до 15 година из 22 школе у Србији. За потребе истраживања креирани су и евалуирани одговарајући психометријски инструменти. Анализирана је веза профила вишеструких интелигенција ученика виших разреда основне школе са склоностима и навикама играња дигиталних игара. Проучавана је могућност предикције префериране врсте дигиталних игара у зависности од профила вишеструких интелигенција и навика играња дигиталних игара. Утврђено је да је профил вишеструких интелигенција валидан предиктор склоности, навика играња и идентификатор зависности од дигиталних игара.Извршена је идентификација зависности од дигиталних игара и анализирана је њена повезаност са профилима вишеструких интелигенција. Анализирани су утицаји појединих социо- демографских фактора на склоности и навике играња дигиталних игара и зависност од дигиталних игара. У складу са циљем и задацима истраживања формиране су хипотезе, које су дискутоване и потврђене. На основу резултата утврђено је постојање везе играња дигиталних игара са профилом вишеструких интелигенција ученика виших разреда основе школе., Disertacija pruža empirijski utemeljen inovativan i jedinstven uvid u vezu igranja digitalnih igara sa profilom višestrukih inteligencija učenika viših razreda osnove škole. Sistematizovan pregled literature realizovan u okviru disertacije ukazuje na činjenicu da se istraživači slažu u stavu da su digitalne igre postale metodički validan obrazovni alat, ali da postoji veoma mali broj empirijskih istraživanja posvećenih vezi između igranja digitalnih igara i profila višestrukih inteligencija učenika. Istraživanje je realizovano na uzorku od 1262 učenika uzrasta od 11 do 15 godina iz 22 škole u Srbiji. Za potrebe istraživanja kreirani su i evaluirani odgovarajući psihometrijski instrumenti. Analizirana je veza profila višestrukih inteligencija učenika viših razreda osnovne škole sa sklonostima i navikama igranja digitalnih igara. Proučavana je mogućnost predikcije preferirane vrste digitalnih igara u zavisnosti od profila višestrukih inteligencija i navika igranja digitalnih igara. Utvrđeno je da je profil višestrukih inteligencija validan prediktor sklonosti, navika igranja i identifikator zavisnosti od digitalnih igara.Izvršena je identifikacija zavisnosti od digitalnih igara i analizirana je njena povezanost sa profilima višestrukih inteligencija. Analizirani su uticaji pojedinih socio- demografskih faktora na sklonosti i navike igranja digitalnih igara i zavisnost od digitalnih igara. U skladu sa ciljem i zadacima istraživanja formirane su hipoteze, koje su diskutovane i potvrđene. Na osnovu rezultata utvrđeno je postojanje veze igranja digitalnih igara sa profilom višestrukih inteligencija učenika viših razreda osnove škole., Dissertation provides an empirically grounded innovative and unique insight into relations between digital gameplay and multiple intelligences profile. A systematic literature review that was realized in this dissertation points to thefact that researchers agree that digital games became methodically valid educational tool, but that there are very few empirical studies on relations between digital gameplay and multiple intelligences profile of the students.The research sample consisted of 1262 11-to-15- year-old students from 22 Serbian schools. The adequate psychometric instruments were developed and evaluated as part of the research. The relations between multiple intelligence profile of higher grade elementary school students and their gameplay preferences and habits were analyzed. The possibility of predicting preferred digital games genre based on multiple intelligences profile and digital gameplay was examined. It was found that the multiple intelligences profile is a valid predictor of digital gameplay preferences, habits and addiction. The digital games addiction was identified, and its relation with the structure of multiple intelligence profile was examined. The influence of socio-demographic factors on digital gameplay preferences, habits and addiction was analyzed. In accordance with the dissertation objectives and tasks the research hypotheses were formulated, and further discussed and confirmed. The results imply the existence of the relation between digital gameplay and multiple intelligences profile of higher grade elementary school students.

Successful prediction of defects at an early stage is one of the main goals of software quality assurance. Having an indicator of the severity of defect occurrence may bring further benefit to allocation of testing resources. Code churn in terms of modified lines of code was found to be a good indicators of bugs. However, that metric does not reveal the structural changes of the code design. The idea of this project is to analyze the relationship between the evolution of object oriented software metrics and the appearance of defects. To achieve this, the absolute and relative differences between the initial version of a class and its current version were calculated for each metrics as indicators of design change. Then, the correlation between these differences and the number of defects were analyzed. Our case study showed that certain metrics have no influence on defect occurrence, while several of them exhibit moderate level of correlation. In addition, we concluded that the relative differences were inappropriate indicator for determining the relationship.

This paper summarizes a novel approach for consuming a domain specific language (DSL) by transforming it to a directed graph representation persisted by a graph database. Using such specialized database enables advanced navigation trough the stored model exposing only relevant subsets of meta-data to different involved services and components. We applied this approach in a mobile cloud computing system and used it to model several mobile applications in retail, supply chain management and merchandising domain. These application are distributed in a Software-as-a-Service (SaaS) fashion and used by thousands of customers in Croatia. We report on lessons learned and propose further research on this topic.

The results of empirical case studies in Software Defect Prediction are dependent on data obtained by mining and linking separate software repositories. These data often suffer from low quality. In order to overcome this problem, we have already investigated all the issues that influence the data collection process, proposed a systematic data collection procedure and evaluated it. The proposed collection procedure is implemented in the Bug-Code Analyzer tool and used on several projects from the Eclipse open source community. In this paper, we perform additional analysis of the collected data quality. We investigate the impact of untraceable bugs on non- fault-prone category of files, which is, to the best of our knowledge, an issue that has never been addressed. Our results reveal this issue should not be an underestimated one and should be reported along with bugs’ linking rate as a measure of dataset quality.

Doktorska disertacija se bavi istraživanjima koja imaju za cilj da unapredeobrazovanje nastavničkog kadra iz oblasti informatike. Predmet istraživanja disertacije jerazvoj platforme koja će olakšati sinhronizaciju razvoja nastavničkog kurikuluma izinformatičke oblasti sa aktuelnim srednjoškolskim informatičkim standardima isavremenim kurikulumima za nastavnike informatike. Prvi pravac istraživanja jeidentifikovanje stanja u oblasti obrazovanja nastavnika informatike (modalitetiobrazovanja nastavnika i kompetencije koje se tim modalitetima stiču). U disertaciji sestanje identifikuje na osnovu analize savremene literature, aktuelnih standarda,reprezentativnih kurikuluma i preliminarnog istraživanja nad populacijom sadašnjihnastavnika informatike na teritoriji Vojvodine. Drugi pravac istraživanja jeobezbeđivanje (tehničke) pretpostavke za objektivnije i efikasnije kreiranje,upoređivanje i razmenu kurikuluma. Taj pravac se odnosi na razvoj mašinski čitljiveontologije kurikuluma bazirane na većpostojećim standardima u domenu obrazovanja isoftverske podrške za upoređivanje takvih reprezentacija kurikuluma bazirane natehnologijama Semantičkog Weba.Ciljevi istraživanja u disertaciji su specifikacija mašinski čitljivog modela informatičkogkurikuluma za obrazovanje nastavnika informatike, mašinski čitljivog modelainformatičkog kurikuluma za srednjoškolski nivo obrazovanja i razvoj prototipasoftverske aplikacije za verifikaciju predloženih modela., The doctoral thesis deals with researches that aim to improve the education ofteaching staff in the field of informatics. The subject of the dissertation research is todevelop a platform that will facilitate the synchronization of the development of teaching curricula in the informatics field with the current secondary informatics standards and modern curricula for informatics teachers of informatics. The first research direction is to identify the state in the informatics teacher education (modalities of teacher education and competences acquired through these modalities). In the dissertation, the state is identified from the review of literature, currentstandards, representative curricula and a preliminary research on population of current informatics teachers in Vojvodina. Another research direction is to provide (technical) prerequisites for more objective and more efficient creation, comparison and exchange of curricula. That direction is related to the development of a machine-readable ontological representation of a curriculum based on existing standards in the field of education and to developing software support for comparing such representations of curricula based on the Semantic Web technologies.The goals of the research in the dissertation are to specify a machine readable model of the informatics teacher education curriculum, a machine-readable model of the informatics curriculum for secondary level education and the development of a software application prototype in order to verify the proposed models.

The subject of this dissertation encompasses a comprehensive overview and analysis of the impact of Sakoe-Chiba global constraint on the most commonly used elastic similarity measures in the field of time-series data mining with a focus on classification accuracy. The choice of similarity measure is one of the most significant aspects of time-series analysis - it should correctly reflect the resemblance between the data presented in the form of time series. Similarity measures represent a critical component of many tasks of mining time series, including: classification, clustering, prediction, anomaly detection, and others. The research covered by this dissertation is oriented on several issues: 1. review of the effects of global constraints on the performance of computing similarity measures, 2. a detailed analysis of the influence of constraining the elastic similarity measures on the accuracy of classical classification techniques, 3. an extensive study of the impact of different weighting schemes on the classification of time series, 4. development of an open source library that integrates the main techniques and methods required for analysis and mining time series, and which is used for the realization of these experiments Predmet istraživanja ove disertacije obuhvata detaljan pregled i analizu uticaja Sakoe-Chiba globalnog ograničenja na najčešće korišćene elastične mere sličnosti u oblasti data mining-a vremenskih serija sa naglaskom na tačnost klasifikacije. Izbor mere sličnosti jedan je od najvažnijih aspekata analize vremenskih serija - ona treba verno reflektovati sličnost između podataka prikazanih u obliku vremenskih serija. Mera sličnosti predstavlјa kritičnu komponentu mnogih zadataka mining-a vremenskih serija, uklјučujući klasifikaciju, grupisanje (eng. clustering), predviđanje, otkrivanje anomalija i drugih. Istraživanje obuhvaćeno ovom disertacijom usmereno je na nekoliko pravaca: 1. pregled efekata globalnih ograničenja na performanse računanja mera sličnosti, 2. detalјna analiza posledice ograničenja elastičnih mera sličnosti na tačnost klasifikacije klasičnih tehnika klasifikacije, 3. opsežna studija uticaj različitih načina računanja težina (eng. weighting scheme) na klasifikaciju vremenskih serija, 4. razvoj biblioteke otvorenog koda (Framework for Analysis and Prediction - FAP) koja će integrisati glavne tehnike i metode potrebne za analizu i mining vremenskih serija i koja je korišćena za realizaciju ovih eksperimenata. Predmet istraživanja ove disertacije obuhvata detaljan pregled i analizu uticaja Sakoe-Chiba globalnog ograničenja na najčešće korišćene elastične mere sličnosti u oblasti data mining-a vremenskih serija sa naglaskom na tačnost klasifikacije. Izbor mere sličnosti jedan je od najvažnijih aspekata analize vremenskih serija - ona treba verno reflektovati sličnost između podataka prikazanih u obliku vremenskih serija. Mera sličnosti predstavlja kritičnu komponentu mnogih zadataka mining-a vremenskih serija, uključujući klasifikaciju, grupisanje (eng. clustering), predviđanje, otkrivanje anomalija i drugih. Istraživanje obuhvaćeno ovom disertacijom usmereno je na nekoliko pravaca: 1. pregled efekata globalnih ograničenja na performanse računanja mera sličnosti, 2. detaljna analiza posledice ograničenja elastičnih mera sličnosti na tačnost klasifikacije klasičnih tehnika klasifikacije, 3. opsežna studija uticaj različitih načina računanja težina (eng. weighting scheme) na klasifikaciju vremenskih serija, 4. razvoj biblioteke otvorenog koda (Framework for Analysis and Prediction - FAP) koja će integrisati glavne tehnike i metode potrebne za analizu i mining vremenskih serija i koja je korišćena za realizaciju ovih eksperimenata.

In modern approach to software development, a great importance is given to monitoring of software quality in early development phases. Therefore, static analysis becomes more important. Furthermore, software projects are becoming more complex and heterogeneous. These characteristics are reflected in a diversity of functionalities and variety of computer languages and the technologies used for their development. Because of that consistency in static analysis becomes more important than it was earlier.In this dissertation SSQSA: Set of Software Quality Static Analyzers is described. The aim of the SSQSA framework is consistent static analysis. This goal is reached by introducing new intermediate source code representation called eCST: enriched Concrete Syntax Tree. The dissertation mostly focuses on eCST, intermediate representations derived from it, and their generation with description of thetools involved in it.The main characteristic of eCST is language independence which gives to SSQSA framework two-level extensibility: supporting a new language and supporting a new analysis. This leads to eciency of adding both level supports andconsistency of added functionalities.To prove the concept, support for more than 10 characteristic languages was introduced. Furthermore, characteristic static analysis techniques (software metrics calculation,code-clone detection, etc.) were implemented and integrated in the framework.Established SSQSA framework provides the infrastructure for the further development of the complete platform for software quality control., U modernim pristupima razvoju softvera veliki značaj pridaje se kontroli kvaliteta softvera u ranim fazama razvoja.Zbog toga, statička analiza postaje sve značajnija. Takođe,softverski proizvodi postaju sve kompleksniji i heterogeni.Ove karakteristike se ogledaju u raznovrsnosti jezika itehnologija koje se koriste u procesu razvoja softvera. Zbogtoga, konzistentnost u statičkoj analizi dobija veći značajnego što je to bio slučaj ranije.U ovoj disertaciji opisan je SSQSA skup statičkih analizatoraza kontrolu kvaliteta (eng. Set of Software Quality StaticAnalyzers). Namena SSQSA okvira je konzistentna statičkaanaliza. Cilj se postiže uvođenjem nove međureprezentacijeizvornog koda nazvane eCST (obogaćeno konkretno sintaksno stablo, eng. enriched Concrete Syntax Tree).Fokus disertacije je primarno na eCST reprezenataciji koda,reprezentacijama izvedenjim iz eCST i procesu njihovoggenerisanja, sa opisom oruđa angažovanim u ovim procesima.Osnovna i najbitnija karakteristika eCST reprezenatacijeje nezavisnost od jezika u kom je izvorni kod pisan, štoSSQSA okviru daje proširivost na dva nivoa: kroz podrškuza nove jezike i kroz podršku za nove analize. Ovo dovodi doefikasnog uvođenja funkcionalnosti na oba navedena nivoa,kao i do konzistentnosti uvedenih funkcionalnosti.Kao dokaz ispravnosti koncepta, podrška za više od 10ulaznih jezika je uvedena. Takođe, implementirane su karakteristične tehnike statičke analize (izračunavanje softverskihmetrika, otkrivanje duplikata u kodu, itd.) i integrisane uSSQSA okvir.Na opisani način, postavljanjem SSQSA okvira, obezbeđenaje infrastruktura za dalji razvoj kompletne platforme za kontrolu kvaliteta softvera.&nbsp

The agent technology represents one of the most consistent approaches to distributed artificial intelligence. Agents are characterized by autonomous, reactive, proactive, and social behavior.In addition, more complex, intelligent agents are often definedin terms of human-like mental attitudes, such as beliefs, desires,and intentions.This thesis deals with software agents and multiagent systemsin several ways. First, it defines a new reasoning architecturefor intelligent agents called Distributed Non-Axiomatic Reasoning System (DNARS). Instead of the popular Belief-Intention-Desire model, it uses Non-Axiomatic Logic, a formalism developed for the domain of articial general intelligence. DNARSis highly-scalable, capable of answering questions and derivingnew knowledge over large knowledge bases, while, at the sametime, concurrently serving large numbers of external clients.Secondly, the thesis proposes a novel agent runtime environmentnamed Siebog. Based on the modern web and enterprise stan-dards, Siebog tries to reduce the gap between the agent technology and industrial applications. Like DNARS, Siebog is adistributed system. Its server side runs on computer clustersand provides advanced functionalities, such as automatic agentload-balancing and fault-tolerance. The client side, on the otherhand, runs inside web browsers, and supports a wide variety ofhardware and software platforms.Finally, Siebog depends on DNARS for deploying agents withunique reasoning capabilities., "null", Agentska tehnologija predstavlja dosledan pristup razvoju distribuirane veštačke inteligencije. Ono što agente izdvaja od ostalih pristupa su autonomno, reaktivino, pro-aktivno, i socijalno ponašanje. Pored toga, kompleksniji, inteligentni agenti se često definišu koristeći ljudske mentalne konstrukcije, kao sto su verovanja, želje i namere.Disertacija se bavi softverskim agentima i multiagentskim sistemima sa nekoliko aspekata. Prvo, definisana je nova arhitektura za rasuđivanje sa primenom u razvoju inteligentnih agenata, nazvana Distribuirani sistem za ne-aksiomatsko rasuđivanje (eng. Distributed Non-Axiomatic Reasoning System) (DNARS). Umesto popularnog BDI modela za razvoj inteligentnih agenata (eng. Belief-Desire-Intention), arhitektura se zasniva na tzv. ne-aksiomatskoj logici, formalizmu razvijenom u domenu veštačke opšte inteligencije. DNARS je skalabilan softverski sistem, sposoban da odgovara na pitanja i da izvodi nove zaključke na osnovu veoma velikih baza znanja, služeći pri tome veliki broj klijenata.Zatim, u disertaciji je predložena nova multiagentska platforma nazvana Siebog. Siebog je zasnovan na modernim standardima za razvoj veb aplikacija, čime pokušava da smanji razliku izmedu multiagentskih sistema i sistema koji se koriste uindustriji. Kao DNARS, i Siebog je distribuiran sistem. Na serverskoj strani, Siebog se izvršava na računarskim klasterima, pružajući napredne funkcionalnosti, poput automatske distribucije agenata i otpornosti na greške. Sa klijentske strane, Siebogse izvršava u veb pretrazivačima i podržava široku lepezu hardverskih i softverskih platformi.Konačno, Siebog se oslanja na DNARS za ravoj agenata sa jedinstvenim sposobnostima za rasuđivanje.

Almost any large-scale system can be viewed as a network that shows interac-tions among entities which are constituent parts of the system. The focus of thisdissertation is on complex networks from three domains: (1) networks extractedfrom source code of computer programs that represent design of software systems,(2) networks extracted from semantic web ontologies that describe the structureof shared and reusable knowledge, and (3) networks extracted from bibliographicrecords that depict collaboration in science. We proposed new methods for theextraction of networks from mentioned domains. Secondly, on several case stud-ies we demonstrated benets of network-based analysis of concrete systems fromthose domains. In contrast to the previous work on the subject, analyses pre-sented in this dissertation are not purely topological, but combine techniques andmetrics developed under the framework of complex network theory with domain-dependent metrics., Skoro svaki kompleksan sistem se može predstaviti mrežom koja opisuje interakcije izmedju entiteta od kojih je sistem komponovan. Fokus ove disertacije je nakompleksnim mrežama iz tri domena: (1) mreže ekstrahovane iz izvornog kodaračunarskih programa koje reprezentuju dizajn softverskih sistema, (2) mreže ekstrahovane iz ontologija semantičkog web-a koje opisuju strukturu deljenog znanjapogodnog za višekratnu upotrebu, i (3) mreže ekstrahovane iz bibliografskih zapisa koje opisuju saradnju istraživača. U okviru disertacije predložene su novemetode za ekstrakciju mreža iz pomenutih domena. Drugo, na nekoliko studijaslučaja ilustrovani su beneti mrežno orjentisane analize konkretnih sistema izdomena obuhvaćenih disertacijom. U poredjenju sa prethodnim relevantim istraživanjima, analize prezentovane u disertaciji nisu čisto topološke, nego kombinuju tehnike i metrike razvijene u okviru teorije kompleksnih mreža sa metrikama iz konkretnog domena.

Predmet istraživanja disertacije obuhvata realizaciju opšteg modela tutorskogsistema za elektronsko učenje iz različitih domena primenom tehnologija semantičkogveba i primena tog modela za izgradnju tutorskog sistema za učenje programskog jezikaJava sa elementima personalizacije.Cilj disertacije je implementacija i predstavljanje svih elemenata tutorskog sistema zaučenje programskog jezika Java pomodu tehnologija semantičkog veba. Ovaj procesobuhvata kreiranje osnovnih gradivnih ontologija kao i pravila za izvođenje konkretnihakcija kojim se postiže personalizacija nastavnog materijala., The subject of the dissertation includes the implementation of a conceptualmodel of tutoring system for e-learning in different domains using semantic webtechnologies and application of that model in a design of a tutoring system forpersonalised learning of Java programming language.The goal of the dissertation is the implementation and presentation of all elements of the tutoring system for learning the Java programming language using semantic webtechnologies. This process includes the creation of the fundamental building blocks ofontologies and rules for carrying out the actions for adaptation of teaching materials.

U stvarnom životu većina informacija dolazi kao neprecizne ilinepotpune vrednosti. XML tehnologija je napravila velikinapredak u oblasti skladištenja i prenosa podataka. Doktorskadisertacija definiše fazi XML sintaksu koja kombinuje neodređenostiu vrednostima XML-a i neodređenosti u strukturi XML dokumenata.Takođe, rad proširuje standardnu XQuery upitnu sintaksu fazielementima i uvodi prioritete i pragove zadovoljenja pomoćuGPFCSP-a. Za razliku od drugih radova koji se fokusiraju napostavljanje teorija i definisanje sintaksi, rad posebnu pažnjuposvećuje praktičnoj upotrebi definisanih sintaksi. Za potrebeistraživanja razvijen je softverski paket koji omogućava rad kako sastandardnim, tako i sa fazi XML, XSD i DTD dokumentima, kao ipostavljanje i izvršavanje prioritizovanih fazi XQuery upita. Alat jetestiran nad primerima iz prakse., In real life, as opposed to virtual, most information comes in theform of imprecise or incomplete values. XML technology hasmade great progress in the field of storage and data transfer.This doctoral dissertation developed XML extension whichcombines indefiniteness in the values of XML andindefiniteness in the structure of XML into a single fuzzy XMLextension. Furthermore, dissertation expands XQuery syntaxwith fuzzy values and includes priorities and thresholds infuzzy XQuery extension using GPFCSP for the first time.Unlike other papers that focus on setting up theories anddefining syntax, in this dissertation, special attention is turnedto their practical use. A tool for working with XML, XSD andDTD documents and prioritized fuzzy XQuery extensionqueries has been developed. The tool has been tested onpractical examples.

Predmet istraživanja disertacije je prikaz koncepata elektronskogučenja, opis potrebe, motivacije i metodologije za obradu i konverzijupostojećih tradicionalnih nastavnih materijala, bez obzira na fizički oblik isadržaj, u savremene i aktuelne elektronske didaktičke objekte (learningobjects). Ideja je da oni budu kreirani tako da mogu biti korišćeni, ako jemoguće i više puta, u sistemima za podršku elektronskom učenju.Drugi osnovni predmet istraživanja je opis, definisanje i praktična proverametodologije za razvoj saradničkih i komunikacionih veština studenata.Pri prelazu iz učionice u oblast elektronskog učenja one se zamenjujuelektronskim aktivnostima, standardnim i uobičajenim delom sistema zaelektronsko učenje. Analizirana je primena klasičnih aktivnosti Web 2.0tehnologije: foruma, chat-sistema, Wiki i blog tehnologije, kao imetodologije za on-line zadavanje, rešavanje, pregledanje i ocenjivanjeindividualnih i timskih zadataka.Konačno, data je sistematizacija metodologija upotrebe elektronskih nastavnihmaterijala i načini opisivanja meta-modelom. Prikazani su neki od postojećih alataza upotrebu u elektronskom učenju, a predložen je model i potencijalno korisnialati za automatizaciju procesa konverzije nastavnih materijala koji bi olakšalikorišćenje u sistemima za podršku elektronskom učenju., The research topic involves description of eLearning concepts,depiction of need, motivation and methodology for processing andconversion of existing traditional teaching materials, regardless of theirphysical form and content, into contemporary and up-to-date electronicteaching material, learning objects. The idea is that these learningobjects are created in such a way that they can be used, if possible alsoreused, in some of the learning management systems.The other primary research topic is the description, definition andpractical testing of methodologies for the development of collaborativeand communicative skills with students. Those skills are, as a part of thetransfer from the classroom to the field of eLearning, substituted withelectronic activities which are the standard and usual part of learningmanagement systems. Within the thesis, application of typical activities ofso-called Web 2.0 technologies: forums, chat-systems, Wiki and blogtechnology, but also methodologies for on-line assigning, solving, audit,and assessment of individual and team assignments.Finally, systematization of various methodologies of digital teachingmaterial use is presented, together with the methods of their depiction bymeta-model. Examples of currently existing tools available for eLearningare presented and a model is suggested, together with the potentiallyuseful tools that might help in automatization of a process of conversionof teaching materials and their use in learning management systems.&nbsp

Diverse research groups have identified that analyzing the bugs-to-code fix links may generate many interesting theories. However, this kind of datasets are usually not easily available from the software development projects. Usually, the bug tracking and the related code fix activities are divided into separate processes and use separated repositories thus complicating the linking of the data. Numerous techniques have been developed but we still lack the empirical data collected with standard data collection procedures. In this paper we examine the effectiveness of the bug linking techniques. In particular we evaluate the proposed technique based on the regular expressions and compare its effectiveness with other known bug linking techniques. Our case study shows that the proposed technique is equally effective as some other already proposed techniques in eliminating the linking bias we identified some differences that may have influence on the linking precision. However, the technique is not addressing data quality issues that may be present in software repositories.

The research topic involves personalization of an e‐learning system based oncollaborative tagging techniques integrated in a recommender system. Collaborative tagging systems allow users to upload their resources, and to label them with arbitrary words, so‐called tags. The systems can be distinguished according to what kind of resources are supported. Besides helping user to organize his or her personal collections, a tag also can be regarded as a user’s personal opinion expression. The increasing number of users providing information about themselves through social tagging activities caused the emergence of tag‐based profilingapproaches, which assume that users expose their preferences for certain contents through tag assignments. Thus, the tagging information can be used to make recommendations. Dissertation research aims to analyze and define an enhanced model to select tags that reveal the preferences and characteristics of users required to generate personalized recommendations. Options on the use of models for personalized tutoring system were also considered. Personalized learning occurs when e‐learning systems make deliberate efforts to design educational experiences that fit the needs, goals, talents, learning styles, interests of theirlearners and learners with similar characteristics. In practice, models defined in the dissertation were evaluated on tutoring system for teaching Java programming language., Predmet istraživanja disertacije obuhvata personalizaciju tutorskih sistema za elektronsko učenje primenom tehnika kolaborativnog tagovanja (collaborative tagging techniques) integrisanih u sisteme za generisanje preporuka (recommender systems). Tagovi, kao oblik meta podataka, predstavljaju proizvoljne ključne reči ili fraze koje korisnik može da upotrebi za označavanje različitih sadržaja. Pored toga što tagovi korisnicima pružaju pomoć u organizaciji sadržaja, oni su korisni i u izražavanju mišljenja korisnika. Veliki broj informacija koje korisnici pružaju o sebi kroz aktivnosti tagovanja otvorio je mogućnost primene tagova u generisanju preporuka. Istraživanje disertacije je usmereno na analizu i definisanje poboljšanih modela za odabir tagova koji otkrivaju sklonosti i osobine korisnika potrebne za generisanje personalizovanih preporuka. Razmatrane su i mogućnosti primene tako dobijenih modela za personalizaciju tutorskih sistema. Personalizovani tutorski sistemi korisniku pružaju optimalne putanje kretanja i adekvatne aktivnosti učenja na osnovu njegovih osobina, njegovog stila učenja, znanja koje on poseduje u toj oblasti, kao i prethodnog iskustva korisnika sistema koji imaju slične karakteristike. Modeli definisani u disertaciji u praksi su evaluirani na tutorskom sistemu za učenje programskog jezika Java.

Software defect prediction is an important decision support activity in software quality assurance. Its goal is reducing verification costs by predicting the system modules that are more likely to contain defects, thus enabling more efficient allocation of resources in verification process. The problem is that there is no widely applicable well performing prediction method. The main reason is in the very nature of software datasets, their imbalance, complexity and properties dependent on the application domain. In this paper we suggest a research strategy for the study of the performance stability using different machine learning methods over different levels of imbalance for software defect prediction datasets. We also provide a preliminary case study on a dataset from the NASA MDP open repository using multivariate binary logistic regression and forward and backward feature selection. Results indicate that the performance becomes unstable around 80% of imbalance.

In this paper we present a framework for developing an intelligent job management and scheduling system that utilizes application specific benchmarks to mould jobs onto available resources. In an attempt to achieve the seemingly irreconcilable goals of maximum usage and minimum turnaround time this research aims to adapt an open-framework benchmarking scheme to supply information to a mouldable job scheduler. In a green IT obsessed world, hardware efficiency and usage of computer systems becomes essential. With an average computer rack consuming between 7 and 25 kW it is essential that resources be utilized in the most optimum way possible. Currently the batch schedulers employed to manage these multi-user multi-application environments are nothing more than match making and service level agreement (SLA) enforcing tools. These management systems rely on user prescribed parameters that can lead to over or under booking of compute resources. System administrators strive to get maximum “usage efficiency” from the systems by manual fine-tuning and restricting queues. Existing mouldable scheduling strategies utilize scalability characteristics, which are inherently 2dimensional and cannot provide predictable scheduling information. In this paper we have considered existing benchmarking schemes and tools, schedulers and scheduling strategies, and elastic computational environments. We are proposing a novel job management system that will extract performance characteristics of an application, with an associated dataset and workload, to devise optimal resource allocations and scheduling decisions. As we move towards an era where on-demand computing becomes the fifth utility, the end product from this research will cope with elastic computational environments.

The proposed theme relates to the field of application of modern computer arithmetics in the analysis of process performance and computational iterations, where the concept of the modern computer arithmetic applies to multiple-precision arithmetic and interval arithmetic involved in the new standard IEEE 754 in 2008. Advance computer arithmetic, first of all interval arithmetic and multi-precision arithmetic, are employed in the the dissertation for the analysis of matrix models in the design iteration process software and iterative numerical computations. These are important and current research topics from the point at which the application is working in the world. Research in this area have led to the development and analysis of algorithms to control the accuracy, optimality, rate calculations and other performance aspects of various processes such as designing software and hardware, designing industrial products, transportation optimization, modeling systems, the implementation of numerical algorithms and others. The first part of dissertation is devoted to engineering design and development of new products, which are either industrial products, technical innovations, hardware or software, often contain a very complex set of relationships among many coupled tasks. Controlling, redesigning and identifying features of these tasks can be usefully performed by a suitable model based on the design structure matrix in an iteration procedure. The proposed interval matrix model of design iteration controls and predicts slow and rapid convergence of iteration work on tasks within a project. A new model is based on Perron- Frobenius theorem and interval linear algebra where intervals and interval matrices are employed instead of real numbers and real matrices. In this way a more relaxed quantitative estimation of tasks is achieved and the presence of undetermined quantities is allowed to a certain extent. The presented model is demonstrated in the example of simplified software development process. An additional contribution in this dissertation is the ranking of tasks within a design mode using components of eigenvalue vector corresponding to the spectral radius of design structure matrix. The second part deals with computational efficiency of numerical iterative algorithms. Computational cost of iterative procedures for the implementation of basic arithmetic operations in multi-precision arithmetic are studied and later applied for the analysis of computational efficiency of the iterative methods for solving nonlinear equations. A new approach that deals with the weights of employed arithmetic operations, involved in the realization of these algorithms, is proposed. This enables a precise ranking of considered root-finding algorithms of different structure, especially in a regime of variable (dynamic) precision of applied multi-precision arithmetic.

It has been estimated that by the end of 2020 there will be 50 billion connected devices in Machine-to-Machine (M2M) networks. Such projections should encourage us to deal with the corresponding problems in heterogeneous M2M systems. First of all, devices can communicate through different access technologies (e.g. wireline, 2G/3G, WiFi, Bluetooth) and their communication can be classified as direct or indirect, internal or external. In this paper we explain differences between those types of communication and propose a new identification scheme that allows M2M devices to establish communication in every possible way. Secondly, there is a problem of device hardware and software diversity. To over-come this problem, we propose the usage of the Open Service Gateway Initiative (OSGi) framework.

In the current information age, massive amounts of data are gathered, at a rate prohibiting their effective structuring, analysis, and conversion into useful knowledge. This information overload is manifested both in large numbers of data objects recorded in data sets, and large numbers of attributes, also known as high dimensionality. This dis-sertation deals with problems originating from high dimensionality of data representation, referred to as the “curse of dimensionality,” in the context of machine learning, data mining, and information retrieval. The described research follows two angles: studying the behavior of (dis)similarity metrics with increasing dimensionality, and exploring feature-selection methods, primarily with regard to document representation schemes for text classification. The main results of the dissertation, relevant to the first research angle, include theoretical insights into the concentration behavior of cosine similarity, and a detailed analysis of the phenomenon of hubness, which refers to the tendency of some points in a data set to become hubs by being in-cluded in unexpectedly many k-nearest neighbor lists of other points. The mechanisms behind the phenomenon are studied in detail, both from a theoretical and empirical perspective, linking hubness with the (intrinsic) dimensionality of data, describing its interaction with the cluster structure of data and the information provided by class la-bels, and demonstrating the interplay of the phenomenon and well known algorithms for classification, semi-supervised learning, clustering, and outlier detection, with special consideration being given to time-series classification and information retrieval. Results pertaining to the second research angle include quantification of the interaction between various transformations of high-dimensional document representations, and feature selection, in the context of text classification. U tekućem „informatičkom dobu“, masivne količine podataka se sakupljaju brzinom koja ne dozvoljava njihovo efektivno strukturiranje, analizu, i pretvaranje u korisno znanje. Ovo zasićenje informacijama se manifestuje kako kroz veliki broj objekata uključenih u skupove podataka, tako i kroz veliki broj atributa, takođe poznat kao velika dimenzionalnost. Disertacija se bavi problemima koji proizilaze iz velike dimenzionalnosti reprezentacije podataka, često nazivanim „prokletstvom dimenzionalnosti“, u kontekstu mašinskog učenja, data mining-a i information retrieval-a. Opisana istraživanja prate dva pravca: izučavanje ponašanja metrika (ne)sličnosti u odnosu na rastuću dimenzionalnost, i proučavanje metoda odabira atributa, prvenstveno u interakciji sa tehnikama reprezentacije dokumenata za klasifikaciju teksta. Centralni rezultati disertacije, relevantni za prvi pravac istraživanja, uključuju teorijske uvide u fenomen koncentracije kosinusne mere sličnosti, i detaljnu analizu fenomena habovitosti koji se odnosi na tendenciju nekih tačaka u skupu podataka da postanu habovi tako što bivaju uvrštene u neočekivano mnogo lista k najbližih suseda ostalih tačaka. Mehanizmi koji pokreću fenomen detaljno su proučeni, kako iz teorijske tako i iz empirijske perspektive. Habovitost je povezana sa (latentnom) dimenzionalnošću podataka, opisana je njena interakcija sa strukturom klastera u podacima i informacijama koje pružaju oznake klasa, i demonstriran je njen efekat na poznate algoritme za klasifikaciju, semi-supervizirano učenje, klastering i detekciju outlier-a, sa posebnim osvrtom na klasifikaciju vremenskih serija i information retrieval. Rezultati koji se odnose na drugi pravac istraživanja uključuju kvantifikaciju interakcije između različitih transformacija višedimenzionalnih reprezentacija dokumenata i odabira atributa, u kontekstu klasifikacije teksta.

This thesis describes one promising approach where a problem of timeseries analysis and prediction was solved by using Case Based Reasoning(CBR) technology. Foundations and main concepts of this technology aredescribed in detail. Furthermore, a detailed study of different approaches intime series analysis is given. System CuBaGe (Curve Base Generator) - Arobust and general architecture for curve representation and indexing timeseries databases, based on Case based reasoning technology, wasdeveloped. Also, a corresponding similarity measure was modelled for agiven kind of curve representation. The presented architecture may beemployed equally well not only in conventional time series (where allvalues are known), but also in some non-standard time series (sparse,vague, non-equidistant). Dealing with the non-standard time series is thehighest advantage of the presented architecture., U ovoj doktorskoj disertaciji prikazan je interesantan i perspektivan pristuprešavanja problema analize i predviđanja vremenskih serija korišćenjemCase Based Reasoning (CBR) tehnologije. Detaljno su opisane osnove iglavni koncepti ove tehnologije. Takođe, data je komparativna analizarazličitih pristupa u analizi vremenskih serija sa posebnim osvrtom napredviđanje. Kao najveći doprinos ove disertacije, prikazan je sistemCuBaGe (Curve Base Generator) u kome je realizovan originalni načinreprezentacije vremenskih serija zajedno sa, takođe originalnom,odgovarajućom merom sličnosti. Robusnost i generalnost sistemailustrovana je realnom primenom u domenu finansijskog predviđanja, gdeje pokazano da sistem jednako dobro funkcioniše sa standardnim, ali i sanekim nestandardnim vremenskim serijama (neodređenim, retkim ineekvidistantnim).

This thesis describes one promising approach where a problem of timeseries analysis and prediction was solved by using Case Based Reasoning(CBR) technology. Foundations and main concepts of this technology aredescribed in detail. Furthermore, a detailed study of different approaches intime series analysis is given. System CuBaGe (Curve Base Generator) - Arobust and general architecture for curve representation and indexing timeseries databases, based on Case based reasoning technology, wasdeveloped. Also, a corresponding similarity measure was modelled for agiven kind of curve representation. The presented architecture may beemployed equally well not only in conventional time series (where allvalues are known), but also in some non-standard time series (sparse,vague, non-equidistant). Dealing with the non-standard time series is thehighest advantage of the presented architecture., U ovoj doktorskoj disertaciji prikazan je interesantan i perspektivan pristuprešavanja problema analize i predviđanja vremenskih serija korišćenjemCase Based Reasoning (CBR) tehnologije. Detaljno su opisane osnove iglavni koncepti ove tehnologije. Takođe, data je komparativna analizarazličitih pristupa u analizi vremenskih serija sa posebnim osvrtom napredviđanje. Kao najveći doprinos ove disertacije, prikazan je sistemCuBaGe (Curve Base Generator) u kome je realizovan originalni načinreprezentacije vremenskih serija zajedno sa, takođe originalnom,odgovarajućom merom sličnosti. Robusnost i generalnost sistemailustrovana je realnom primenom u domenu finansijskog predviđanja, gdeje pokazano da sistem jednako dobro funkcioniše sa standardnim, ali i sanekim nestandardnim vremenskim serijama (neodređenim, retkim ineekvidistantnim).