|
|
Automatic Keyword Detection
Mašláňová, Marcela ; Karafiát, Martin (referee) ; Smrž, Pavel (advisor)
The main goal of this work is to survey the field of the automatic keywords tagging in a text and apply this background for automatically generating back-of-the-book indexes. Human made indexes are expensive and that's why we are looking for (semi)-automatic methods indexes. The theoretical part of this thesis deals with collocations, which are an important part of generated indexes. The practical part of the work applies selected methods to testing data and summarize results of experiments.
|
|
|
Automatic Photography Categorization
Veľas, Martin ; Řezníček, Ivo (referee) ; Španěl, Michal (advisor)
This thesis deals with content based automatic photo categorization. The aim of the work is to experiment with advanced techniques of image represenatation and to create a classifier which is able to process large image dataset with sufficient accuracy and computation speed. A traditional solution based on using visual codebooks is enhanced by computing color features, soft assignment of visual words to extracted feature vectors, usage of image segmentation in process of visual codebook creation and dividing picture into cells. These cells are processed separately. Linear SVM classifier with explicit data embeding is used for its efficiency. Finally, results of experiments with above mentioned techniques of the image categorization are discussed.
|
|
|
Application of virtual local area networks for effective and flexible data network design
Brázda, Libor ; Burda, Karel (referee) ; Novotný, Vít (advisor)
This thesis is aimed to the problems of virtual local networks. In the first chapters are explained both basic and advanced methods of designing and management of virtual LAN networks. The following chapters are dedicated to the methods of effective management VLAN networks and to the protocol Spanning Tree, preventing formation of loops in redundant networks, and their exploitation in VLANs. The practical part describes the application of the implementation of the technology of virtual local networks in experimental network, containing various types of devices. Based on this theory, a laboratory assignment has been designed for the training of the students in solving such problems. The scope of the laboratory assignment is divided into simpler partial tasks, which even a less advanced student should be able to configure and test.
|
|
|
The potential use of PIT telemetry for identifying and tracking Spinycheek crayfish in their natural environment.
VÍCH, Petr
Main purpose of this diploma work was to find out possibility using PIT telemetry in monitoring strange crayfish species Spinycheek crayfish Orconectes Limosus. Research of the theme consisted of three steps: two laboratory researches and one research in the natural environment. The first laboratory study concerned tagging passive integrated transponder (PIT) to various part of crayfish body. PIT tags were internally implanted to cephalothorax, abdomen and claw. The highest mortality was in tagging to claw. The second laboratory research consisted of comparing different tagging methods to PIT telemetry. Crayfish were tagging by PIT telemetry, soldering, visible implant elastomers and visible implant Alpha tags. Among tagging groups there were not to find out different in mortality. In the third research the crayfish were tagging in their natural environment. There were using three methods: PIT telemetry, soldering and VI Alpha tags. This experiment wasn´t successful. During recapturing crayfish there were not to find any tagging crayfish.
|
|
| |
|
| |
|
|
Dva typy značkování v češtině
Spoustová, D. ; Hajič, J. ; Votrubec, J. ; Krbec, P. ; Květoň, Pavel
Description of several hybrid disambiguation methods combining the strength of hand-written disambiguation rules and statistical taggers. Three different statistical (HMM, Maximum-Entropy and Averaged Perceptron)taggers are used in a tagging experiment using Prague Dependency Treebank. The results of the hybrid systems are better than any other method tried for Czech tagging so far.
|
|
|
Od korpusu jako otevřeného zdroje pro bádání ke komerčním produktům
Šimandl, Josef
The development of corpora is sketched, from large collections of texts without tagging through tagged corpora to machines that operate above tagged corpora and produce data presented as data about language, such as Word Sketches (TM). The article remarks that every corpus is merely a representation of texts and that the quality of representation is to be examined. The unavoidable question in research is how is the corpus built and how, under what principles, the service software operates. Both in case we explore a corpus with distortions, where texts appear in a way nobody has written them so (digits and their environment uses to be phenomena of that sort), and in case we are not allowed to have an insight "below the bonnet" or to change working parameters, we hardly may speak about doing scholarly research.
|
|
| |
|
|
Social Bookmarking
Holásek, Jakub ; Rauch, Jan (advisor) ; Svátek, Vojtěch (referee) ; Nováček, Jan (referee)
Práce je základním informačním podkladem pro pochopení Social Bookmarkingu, jeho aspektů, nástrojů a využití nejen pro jednotlivé uživatele ale i pro skupiny. Zahrnuje zhodnocení těchto nástrojů, jejich kladné i záporné stránky včetně praktických ukázek.
|
guest ::
