Search

Cilj diplomskega dela je razvoj mobilne rešitve z uporabo različnih programskih ogrodij. Prvi del diplomskega dela je namenjen predstavitvi uporabljenih tehnologij. Drugi in najbolj obsežni del diplomskega dela vsebuje predstavitev razvoja mobilne aplikacije in je razdeljen na tri dele. Prvi del vsebuje opis metodologije in ideje razvoja aplikacije ter analize, kjer določimo zahteve in specifikacije zahtev mobilne aplikacije. Drugi del je namenjen opisu razvoja mobilne aplikacije v avtohtonem ogrodju in tretji del je namenjen opisu razvoja mobilne aplikacije v PhoneGap oziroma hibridnem ogrodju. V tretjem delu diplomskega dela pa predstavimo primerjavo med avtohtonim razvojnim ogrodjem in Phonegap oziroma hibridnim razvojnim ogrodjem. The following Bechelor’s thesis discusses the development of mobile application using various programming frameworks. The first part of the thesis is devoted to the presentation of the technologies and tools used in mobile application development. The second and most extensive part of the thesis presents the development of mobile applications and is divided into three parts. The first part describes methodology, application development idea and analysis which sets out the requirements and specifications of the requirements for mobile applications. The second part is dedicated to the description of the development of mobile application with native framework and the third part is dedicated to the description of the development of mobile application with hybrid framework. The third part of the thesis discusses a comparison between native and hybrid frameworks.

Pri inteligentnih tutorskih sistemih je pomembno, da sistem razume, kako težak je določen problem za učenca. Kako samodejno oceniti tovrstno težavnost, ostaja odprto vprašanje. Namen raziskave je razviti algoritmičen pristop k ugotavljanju težavnosti, ki bi ga lahko uporabljali pri avtomatiziranem ocenjevanju težavnosti problemov za človeka. Osredotočili se bomo na ocenjevanje težavnosti problemov, pri katerih težavnost izvira iz kombinatorične kompleksnosti in kjer je potrebno preiskovanje med alternativami. Pristop temelji na uporabi hevrističnega računalniškega preiskovanja za gradnjo preiskovalnih dreves, ki so ``smiselna'' z vidika osebe, ki problem rešuje. Pokazali bomo, da je s pomočjo analize lastnosti tovrstnih ``smiselnih'' dreves računalniški program sposoben napovedati, kako težak za reševanje je določen problem. Naš program je bil sposoben z visoko stopnjo natančnosti ločevati med enostavnimi in težkimi šahovskimi taktičnimi problemi. In intelligent tutoring systems, it is important for the system to understand how difficult a problem is for the student. However, it is an open question how to automatically assess such difficulty. The aim of our research is to find formalized measures of difficulty that could be used in automated assessment of the difficulty of a mental task for a human. We present a computational approach to estimating the difficulty of problems in which the difficulty arises from the combinatorial complexity of problems where a search among alternatives is required. Our approach is based on a computer heuristic search for building search trees that are “meaningful” from a human's point of view. We demonstrate that by analyzing properties of such trees, the program is capable to predict how difficult it would be for a human to solve the problem. In the experiments with chess tactical problems our program was able to differentiate between easy and difficult problems with a high level of accuracy.

BI4Dynamics je analitična BI-rešitev, ki nadgradi transakcijske sisteme Microsoft Dynamics ERP s sposobnostjo kompleksnih analiz podatkov. Podatkovno skladišče BI4Dynamics, ki se namesti na Microsoft SQL-strežnik, je zgrajeno po Kimballovem principu zvezdne sheme in predstavlja podatkovni vir za nabor OLAP-kock v analitični bazi. Vendar prava moč rešitve BI4Dynamics leži v njeni platformni zasnovi. Ker v celoti uporablja Microsoftov arhitekturni sklad, ki temelji na analitični in SQL-tehnologiji, v polni meri omogoča spreminjanje katere koli od komponent modula. Ta naloga opisuje razvoj BI4Profiler performančnega modula, ki pa poleg osnovne funkcionalnosti modula, se pravi svoje OLAP-kocke, nabora dimenzij ter pripadajoče komponente v podatkovnem skladišču, vsebuje še dodatno funkcionalnost za zajem statistik, ki smo jo morali implementirati v ogrodje BI4Dynamics. To doseže z uporabo sistemskih pogledov ali DMV-jev (Dynamics Management Views), ki jih ponuja Microsoftov SQL-strežnik. Moja naloga v tem procesu je bila načrtovanje arhitekture te funkcionalnosti in nadzor razvoja modula. Ker smo hoteli zagotoviti neodvisnost in obstojnost teh statističnih podatkov tudi v primeru, ko uporabnik na novo namesti podatkovno skladišče, smo za modul uvedli novo ločeno podatkovno skladišče in analitično bazo, ki pa se namestita samo ob prvi namestitvi originalnega podatkovnega skladišča in se nato ne spreminjata. Podatki v njih pa se obdelajo po tem, ko se obdelajo vsi podatki v originalnem podatkovnem skladišču in analitični bazi. BI4Dynamics is an analytical Business Intelligence (BI) solution for Microsoft Dynamics ERPs. The solution deploys a data warehouse on a Microsoft SQL server and creates an analytical database on a Microsoft Analysis server accordingly, due to SQL’s flexibility, it offers an excellent platform for developing custom BI modules OLAP cubes and dimensions. This paper describes how the BI4Profiler, which is a performace module for BI4Dynamics platform, was developed. In this process my task was to design the core architecture of the module and oversee project development. The BI4Profiler module grew out of our partners’ and our own need for a tool when benchmarking and optimizing custom developed SQL scripts. We wanted to use the existing framework so this tool was packed as a module, thus it can be analyzed using any BI compatible front end tool. To ensure independency of statistic data, the BI4Profiler deploys its own separate data warehouse with staging and presentation area, as well as an OLAP cube with a set of dimensions. Although module, it has an additional core functionality that we had to integrate into the existing framework. During the execution of the scripts, this separate thread pools the SQL Server using DMVs (Dynamics Management Views) for statistics, which first saves said statistics to the data warehouse, thence builds and processes the analytical database.

Potisna tehnologija je mehanizem za pošiljanje podatkov s spletnega strežnika v spletni brskalnik brez zahteve po prenosu s strani odjemalca. Zaradi naraščajoče potrebe po komunikaciji v realnem času in čim boljši uporabniški izkušnji vedno več produkcijskih sistemov vidi prednosti v takšnem načinu podajanja informacij. V okviru diplomske naloge je bil opravljen pregled programskih rešitev za realizacijo potisne tehnologije. Na podlagi zaključkov je bila izbrana najustreznejša programska oprema in izdelan prototip, ki bo služil kot nadgradnja obstoječega sistema za naročanje in upravljanje spletnih storitev računalništva v oblaku. Push technology is a mechanism for sending data from a web server to a web client, where the client makes no explicit requests for the transfer. Due to the increasing need for real-time communication and optimization of the user experience there is a rising number of production systems that seize the advantage of this kind of information transfer. In the thesis we reviewed the existing software solutions for the realization of push technology. Based on the conclusions, we have selected the most appropriate software and developed a prototype that will serve as an upgrade to the existing system for ordering cloud services.

Izobraževalne simulacije se vedno bolj širijo iz tradicionalnih področij in postajajo pomembno orodje za izobraževanje. Čeprav so to zelo praktična orodja za pomoč v klasičnem izobraževanju, so bila do sedaj podprta predvsem v namiznem okolju in se niso tako uveljavila za širšo uporabo. Iz tega razloga smo razvili orodje za predstavitev fizikalnih simulacij v obliki mobilne aplikacije. Aplikacija je podprta s strani različnih naprav, tudi pametnih telefonov in tablic. Deluje na različnih velikostih zaslona in izkorišča prednosti prenosljivosti in računskih zmožnosti naprav. Z uporabo okolja jQuery Mobile, ki je prilagojen za naprave na dotik, smo dosegli, da aplikacija ni uporabna le v namiznem okolju, ampak dostopna tudi na drugih napravah. Razvili smo tudi podporo za hkratno manipuliranje simulacije s strani več uporabnikov v realnem času. V ta namen smo implementirali spletni vtičnik, ki nam omogoča hkratno kreiranje in manipuliranje simulacije na več napravah. Educational simulations are breaking out from traditional areas of use and emerging as an increasingly important tool for education. Even though, they are very practical tool for assisting in the classrooms, they have been supported mostly on desktop environments and didn’t fulfill the desired adoption. For this reason we created the mobile application that is a tool for presenting physics simulations. It is supported by different devices, including mobile phones and tablets, taking advantage of the different display size, and computational power characteristics of the devices. To make this application reachable not just from desktop computers, we used jQuery Mobile, framework that is optimized for touch devices. We have also provided the multiuser support of the simulations, offering users the possibility to manipulate shared real-time simulations. For that purpose we implemented WebSockets, technology that is facilitating live content and the creation of real-time simulations.

Grupiranje v klike je proces združevanja vozlišč v gruče, za katere velja, da so vsa vozlišča med seboj povezana. V sprotnem (on-line) združevanju celoten graf ni znan vnaprej, ampak je na voljo po eno vozlišče naenkrat. Tista vozlišča, ki so že pridružena gruči, ne morejo biti prestavljena v drugo gručo. Naloga je poiskati takšno razvrstitev vozlišč, ki se od optimalne razvrstitve razlikuje čim manj. V tej diplomski nalogi podamo konstantno zgornjo mejo in algoritem (Lazy) za problem sprotnega združevanja v klike, kjer je cilj poiskati razvrstitev vozlišč s čim več povezavami znotraj gruč (problem Max-ECP). Poleg tega podamo ujemajoči zgornji in spodnji meji za problem sprotnega združevanja v klike, kjer je cilj poiskati razvrstitev s čim manj povezavami med gručami (problem Min-ECP). Za oba problema pokažemo, da naraven (Greedy) pristop vodi k linearni rešitvi. Naša metoda Lazy nudi konstantno tekmovalno razmerje, kar se znatno odraža na grafih z veliko vozlišči. Clique clustering is the problem of partitioning a graph into cliques so that some objective function is optimised. In online clustering the input graph is given one vertex at a time, and vertices that have been previously clustered are not allowed to be separated. The objective is to maintain a clustering that never deviates too far from the optimal offline solution. We give a constant competitive upper bound and a strategy (Lazy) for online clique clustering, where the objective function is to maximise the number of edges inside the clusters (Max-ECP). We also give almost matching upper and lower bounds on the competitive ratio for online clique clustering, where we want to minimise the number of edges between clusters (Min-ECP). In addition, we prove that the greedy method only gives linear competitive ratio for these problems. The research result shows that the proposed constant competitive strategy performs significantly better on bigger graphs than the greedy method.

In this thesis we explore the problem of detection of research subdisciplines of a chosen science, based only on the data about citing papers from a chosen batch of papers relevant to the selected science. We directed our attention to the field of Psychology. It is an interesting scientific discipline, with variety of research topics and numerous scientific publications throughout the years. Due to lack of freely accessible centralized database of psychological papers and their relevant citations, the first step of our thesis was collection of papers and their applicable citations. Data was presented in form of a citation network. Citation networks are acyclic directed information networks, where the structure of the network reflects the structure of the information stored in the network vertices. The process of differentiation of research disciplines in the network was performed by applying a state-of-the-art algorithm for community detection. Our choice was the Louvain method, known for its simplicity, effectiveness and speed. Part of the mechanism for rating the detected communities was to name them and examine their connections. Due to the vast quantity of available data and the unfamiliarity with the psychological field, we named the communities based on the measures for cosine similarity between our initially collected psychological papers and the relevant texts for each of the APA divisions of Psychology. Results obtained by the network analysis and the method for community detection are positive. However, the nature of data collection and the influence of our subjective judgment for community naming offer lots of opportunities for further improvement.

In the field of artificial intelligence, the development of driverless cars has advanced to the point where an autonomous car is issued a license. The technology implemented in these driverless cars is proven to be reliable and robust. Bicyclists on the other hand do not benefit from advancements in this technology yet. We tried to develop a warning system for bicyclists using OpenCV. Our main goal was to warn bicyclists of approaching and overtaking cars. The system was developed for smart phone platforms so we were constrained by the battery lifetime. Our final solution still has some flaws but it is a good starting point for future development, so we propose possible improvements.

The Granata research site was established for the study of structure, growth and regeneration of silver fir-European beech single stem selection forest. Three phytocoenological relevés were carried out, all trees ?5 cm diameter at breast height (d.b.h.) were measured by full callipering, tree growth was analysed for sample trees, regeneration of tree species was registered on 63 sampling plots. In addition, historical data from past forest inventories werestudied. In the period from 1952 to 2003, the share of silver fir has decreased from 86% to 26%, the share of Norway spruce increased significantly,whereas the share of beech and sycamore has slightly increased. The current growing stock amounts to 350 m3 ha-1, with large-size diameter trees (d.b.h. ?50 cm) representing 51% of total growing stock. The site is divided into 16 stand patches with significant differences in tree species composition and diameter distribution. Regeneration is sufficient with silver fir prevailing in total number of seedlings (42%). 23% of all seedlings are damaged as a result of game browsing. Granata (GGE Logatec, odd. 22) je raziskovalni objekt za študij prebiralnega gozda na rastišču Omphalodo-Fagetum. V objektu so bili opravljeni fitocenološki popisi, sestoje smo izmerili s polno izmero, analizirali smo priraščanje dreves in pomlajevanje drevesnih vrst. Na podlagi zgodovinskih virov smo analizirali razvoj gozdnih sestojev v obdobju 1952-2003. V tem času se je delež jelke v sestojih zmanjšal s 86 % na 26 %, znatno se je povečal delež smreke, opazen je tudi delež bukve in gorskega javorja. Lesna zaloga je 350 m3 ha-1, delež debelega drevja (d.b.h. ? 50 cm) predstavlja 51 % celotne zaloge. Objekt je razdeljen na 16 sestojev, med njimi so opazne razlike v drevesni sestavi in debelinski zgradbi. Pomlajevanje je ustrezno, v pomladku prevladuje jelka (42 %), 23 % vseh mladic v pomladku je poškodovanih zaradi objedanja divjadi.

Delo obsega opis implementacije sistema računalniškega vida za problem, ki bi ga lahko šteli med industrijske probleme računalniškega vida, to je problem detekcije številke, na katero je padla kroglica pri igralniški ruleti. Opisani so algoritmi, ki so bili razviti za implementacijo tega specialnega sistema. Podani so tudi rezultati testiranja opisanega sistema nad bazo 42-ih posnetkov, ki pokažejo šibke točke sistema. Predlagane so tudi smernice za nadaljnji razvoj opisanega sistema.

Tehnologija se dandanes razvija z neverjetno hitrostjo in hkrati s svojimi rešitvami generira tako nove probleme kot izzive. Pri reševanju izziva avtomatskega modeliranja človeškega obraza iz posamezne frontalne slike, se tako soočamo s problemi na področju avtomatskega iskanja obrazov in njegovih značilk, kot tudi s problemom modeliranja na podlagi parametrov in mapiranja ter teksturiranja frontalne slike na 3D model, ki ga lahko kasneje animiramo. V diplomski nalogi bomo pregledali različne tehnike za iskanje obrazov in njegovih značilk. Preizkusili bomo dve iskalni metodi. Prva deluje v barvnem prostoru RGB, druga pa deluje na osnovi kromatskega barvnega prostora. Na podlagi ugotovitev bomo predlagali lastno metodo s katero dobimo dokaj natančne pozicije in velikosti značilk. Metoda deluje v akromatskem barvnem prostoru, kar nam daje neodvisnost od kromatskih barv, hkrati pa zaznava samo razlike v intenziteti svetlobe, tako da odvisnost od osvetlitve omili tudi pri neenakomerni osvetlitvi. Obrazne značilke odkriva na osnovi znanja, kar je hkrati tudi slabost te metode, saj uporabljena pravila dopuščajo napačne detekcije. Rezultate - odkrite značilke bomo nato uporabili kot parametre pri modeliranju, mapiranju in teksturiranju generičnega 3D modela obraza v programu Maya. Tak model se lahko kasneje animira ročno ali preko različnih vhodnih naprav (tudi v realnem času), kar je lahko osnova za nadaljno delo.

This work presents reasons for digitizing audiovisual archives as well as ways of doing so. It defines terms such as audiovisual work, digital audiovisual archive and layers in audiovisual archives. The process of digitizing an audiovisual work is analyzed. Five case studies are included. The sixth is the main case of documenting theatre plays on video. Technical solutions for digitizing the archive are presented, costs evaluated, storage and filing suggested. The archive may someday become a vital part of a multimedia theatre play database. Such a database is analysed and new options and techniques enabled by audio and video documenting of theatre plays are researched.

Los recursos pesqueros de la cuenca del Salado poseen una extraordinaria importancia en el contexto de las aguas continentales de la Argentina. Sin embargo, la diversidad tipológica en la cuenca, así como la ausencia de continuidad en la planificación regional, han dificultado una gestión adecuada de los recursos pesqueros. Este trabajo presenta una visión general de los principales aspectos relacionados con la ictiofauna de la región, con un enfoque natural de los procesos y mecanismos de gestión. Se incluye una descripción de las comunidades de peces de la cuenca, identificando aquellas especies de mayor interés deportivo y comercial. Se describen las distintas artes y técnicas de pesca utilizadas en la provincia para la pesca deportiva, comercial y científica. Se revisan los criterios de diagnosis de las poblaciones de pejerrey como principal recurso de interés pesquero y bajo esta perspectiva se proponen nuevos enfoques dirigidos a promover un manejo adecuado de los recursos y su uso sustentable. Se identifican los distintos tipos de pesquerías de practica usual en la cuenca y realiza un análisis de los aspectos socioeconómicos relacionados. Así mismo, se plantea el desarrollo de un nuevo marco de gestión institucional y regulatorio dirigido a optimizar el manejo y la gestión de los recursos pesqueros. Por ultimo, se definen criterios para la ordenación y conservación de dichos recursos y se identifican ejes de conflicto y requerimientos para su aprovechamiento sustentable en el marco de nuevas propuestas para la formulación de políticas publicas., Facultad de Ciencias Naturales y Museo

This work deals with cell nucleus segmentation and the edge detection of the clumps formed by nuclei. The fact is that it is both cell nucleus features and clump shapes that serve as cancer diagnosis criteria. In addition, the work includes several procedures of cell nucleus segmentation described in publications on computing. The author has developed a method for the edge detection of nucleus clumps. It is based on grouping techniques with the greedy algorithm, relaxation and the graph search. The analysis is followed by some examples showing the use of the edge detection method. Differences between normal and malignant ducts are shown on the basis of some pictures of mammary glands. Finally, some improvements and future work extensions are suggested.

Naloga obravnava analizo koron prstov s pomočjo tehnike Gas Discharge Visualization (GDV), Kirlianove kamere in z njo povezane programske opreme. Na primeru študije ugotavljanja efekta različnih majic na posnetke koron prstov s Kirlianovo kamero poskušamo podpreti področje kirlianografije z metodami strojnega učenja in statistike. V okviru te študije je bilo posnetih 187 ljudi, ki so nosili štiri različne majice ali pa so bili v kontrolni skupini. Rezultati študije kažejo pomembne razlike med zdravilnimi in navadnimi majicami ter kontrolno skupino. Naloga nadalje pokaže, da GDV posnetki koron prstov vsebujejo koristne informacije o zdravstvenem stanju človeka. Opisane so ugotovitve glede vpliva nekaterih dejavnikov na GDV posnetke in ugotovitve o medsebojni povezanosti nekaterih GDV parametrov. Naloga vsebuje tudi nasvete za snemanje s kamero, delo z ljudmi in nasvete o tem, kako in katere GDV parametre je smiselno uporabljati pri različnih vrstah analize koron prstov. Dotakne se še problema, kako upoštevati različne pomembnosti sprememb pri GDV parametrih in nakaže nekaj možnih rešitev.