|
|
Semantic Similarity Methods in Folksonomies
Kadlec, Jan ; Otrusina, Lubomír (referee) ; Schmidt, Marek (advisor)
Bakalářská práce byla vypracována na studijním pobytu na "Aalborg University" v Dánsku, a byla zpracována v angličtině. Folksonomie jsou nový, uživateli řízený přístup ke klasifikaci a důležitá čast Web 2.0. Jsou také jediným přístupem, který je schopen udržet krok s dnešní rychlostí expanze webu, tím že předá uživatelům odpovědnost za klasifikaci. Pokud folksonomie obsahují dostatečné množství dat, dají se k mnohému využít. Tato práce se zaměřuje na metody sémantické podobnosti ve folksonomiích. Cílem této práce bylo odzkoušet mnohé metody na vzorku tří datových sad - delicious.com, Last.fm a medworm.com. Toto bylo vykonáno za pomocí kotvících dat z WordNetu, Open Directory Project a zdravotně orientované ontologie. Výsledky přinesené touto prací indikují, že metody sémantické podobnosti mohou být použity k úspěšnému měření podobností v mnohých doménách.
|
|
|
Automatic Photography Categorization
Veľas, Martin ; Beran, Vítězslav (referee) ; Španěl, Michal (advisor)
This thesis deals with content based automatic photo categorization. The aim of the work is to create an application, which is would be able to achieve sufficient precision and computation speed of categorization. Basic solution involves detection of interesting points, extraction of feature vectors, creation of visual codebook by clustering, using k-means algorithm and representing visual codebook by k-dimensional tree. Photography is represented by bag of words - histogram of presence of visual words in a particular photo. Support vector machines (SVM) was used in role of classifier. Afterwards the basic solution is enhanced by dividing picture into cells, which are processed separately, computing color correlograms for advanced image description, extraction of feature vectors in opponent color space and soft assignment of visual words to extracted feature vectors. The end of this thesis concerns to experiments of of above mentioned techniques and evaluation of the results of image categorization on their usage.
|
|
|
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.
|
|
|
Description of Old Czech Common Nouns Declension (with regard to Automatic Morphological Analysis of Texts in Old Czech Text Bank)
Synková, Pavlína ; Oliva, Karel (advisor) ; Petkevič, Vladimír (referee) ; Vepřek, Miroslav (referee)
The thesis aims at explicit description of Old Czech common nouns declension with regard to its application in a tool for automatic morphological analysis of (digitized) texts in Old Czech. This means that this description is intended to serve as a basis for automatic generation of word forms (jointly with their appropriate morphological information and lemma) which will then be used for assigning morphological categories (gender, number, case) and lemma to word forms occurring in Old Czech digitized texts. The thesis thus develops a base for the first step in transformation of text banks (which currently exist for the Old Czech period) into an Old Czech corpus offering more possibilities for linguistic research. The Old Czech period is defined as a period from the beginning of the 14th century (more precisely from the period when first coherent texts written in Czech appeared) approx. to the end of the 15th century. Nouns were chosen for this work, because they cover approx. 30% of texts in current Czech (which is the highest percentage from all parts of speech). Old Czech texts are taken into account only in a transcribed form (based on transcription rules used in the Old Czech Text Bank developed at the Institute of the Czech Language of the Academy of Sciences of the Czech Republic). On the one...
|
|
|
Tag phenomenon on Flickr.com and its impact on interpretion extension of photography
Janda, Filip ; Láb, Filip (advisor) ; Turek, Pavel (referee)
I Abstract The first chapter of the thesis is devoted to the Flickr.com website and its evolution from its foundation until the present and to the posibility of its usage as a news source. The second chapter further explores the concept of folksonomy itself, it presents its definition, the basic approaches to its typology and singly it disserts on tag, on its forms and functions on the Flickr.com website. In the third chapter we will elaborate on the status of photography in its digital form, on its unstability and predisposition to be subject to language. The forth chapter puts forward the characteristics of text in the sphere of Internet, that of its nonlinearity, and it goes on about titles, captions and text as such in relation to photography. The fifth chapter then details the ability of tags to influence the perception of photography in the cyberspace and it addresses another phenomenon, that of the notes. The sixth and the last chapter presents the results of the "Flickrdream" research, whose objective was to find out to what degree are the tags in relation with photos on Flickr capable to comply with representation of such a complex phenomenon as a dream.
|
|
|
Semantic Similarity Methods in Folksonomies
Kadlec, Jan ; Otrusina, Lubomír (referee) ; Schmidt, Marek (advisor)
Bakalářská práce byla vypracována na studijním pobytu na "Aalborg University" v Dánsku, a byla zpracována v angličtině. Folksonomie jsou nový, uživateli řízený přístup ke klasifikaci a důležitá čast Web 2.0. Jsou také jediným přístupem, který je schopen udržet krok s dnešní rychlostí expanze webu, tím že předá uživatelům odpovědnost za klasifikaci. Pokud folksonomie obsahují dostatečné množství dat, dají se k mnohému využít. Tato práce se zaměřuje na metody sémantické podobnosti ve folksonomiích. Cílem této práce bylo odzkoušet mnohé metody na vzorku tří datových sad - delicious.com, Last.fm a medworm.com. Toto bylo vykonáno za pomocí kotvících dat z WordNetu, Open Directory Project a zdravotně orientované ontologie. Výsledky přinesené touto prací indikují, že metody sémantické podobnosti mohou být použity k úspěšnému měření podobností v mnohých doménách.
|
|
|
Automatic Photography Categorization
Veľas, Martin ; Beran, Vítězslav (referee) ; Španěl, Michal (advisor)
This thesis deals with content based automatic photo categorization. The aim of the work is to create an application, which is would be able to achieve sufficient precision and computation speed of categorization. Basic solution involves detection of interesting points, extraction of feature vectors, creation of visual codebook by clustering, using k-means algorithm and representing visual codebook by k-dimensional tree. Photography is represented by bag of words - histogram of presence of visual words in a particular photo. Support vector machines (SVM) was used in role of classifier. Afterwards the basic solution is enhanced by dividing picture into cells, which are processed separately, computing color correlograms for advanced image description, extraction of feature vectors in opponent color space and soft assignment of visual words to extracted feature vectors. The end of this thesis concerns to experiments of of above mentioned techniques and evaluation of the results of image categorization on their usage.
|
|
|
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.
|
guest ::
