This master thesis is oriented on a problem of creating video and image annotations. This problem is solved by crowdsourcing approach. Crowdsourcing games were designed and implemented to make solution of this problem . It was proven by testing that these games are capable of creating high quality annotations. Launching these games on a larger scale could create large database of annotated videos and images.

Recognition and comparison of image is one of the main problems and area of the field of computer vision. This thesis adds to these two issues the third, the recognition image semantics, so called annotations or labels. This work uses the knowledge of methods of recognizing the similarity of images to create a tool that is able based on training dataset of images and annotations, create a group most likely annotation for the test set of images. This work presents several types of test datasets suitable for the detection of annotation information for images. Subsequently, best set with the necessary training dataset size and enough information about annotations is selected. Based on this training dataset algorithm is designed for easy loading test set without large demands on computer performance. Evaluation of annotation information is done based on different similarity algorithms. At the beginning of this work was to use a simple, but not very effective method of MSE and comparison of color histograms, but gradually it was necessary to move to using more advanced methods (such as Tamura, Gabor, CEDD nebo různé druhy hostistogramů). The results of this comparison are then taken to evaluate the likelihood of the annotation for the image specified test set. The last part is an evaluation of the accuracy of annotation based on information from the test set.

Thesis Abstract Akshay Aggarwal July 2020 This thesis attempts at correction of some errors and inconsistencies in dif- ferent treebanks. The inconsistencies can be related to linguistic constructions, failure of the guidelines of annotation, failure to understand the guidelines on annotator's part, or random errors caused by annotators, among others. We propose a metric to attest the POS annotation consistency of different tree- banks in the same language, when the annotation guidelines remain the same. We offer solutions to some previously identified inconsistencies in the scope of the Universal Dependencies Project, and check the viability of a proposed in- consistency detection tool in a low-resource setting. The solutions discussed in the thesis are language-neutral, intended to work with multiple languages with efficiency. 1

Title: Measures of Machine Translation Quality Author: Matouš Macháček Department: Institute of Formal and Applied Linguistics Supervisor: RNDr. Ondřej Bojar, Ph.D. Abstract: We explore both manual and automatic methods of machine trans- lation evaluation. We propose a manual evaluation method in which anno- tators rank only translations of short segments instead of whole sentences. This results in easier and more efficient annotation. We have conducted an annotation experiment and evaluated a set of MT systems using this method. The obtained results are very close to the official WMT14 evaluation results. We also use the collected database of annotations to automatically evalu- ate new, unseen systems and to tune parameters of a statistical machine translation system. The evaluation of unseen systems, however, does not work and we analyze the reasons. To explore the automatic methods, we organized Metrics Shared Task held during the Workshop of Statistical Ma- chine Translation in years 2013 and 2014. We report the results of the last shared task, discuss various metaevaluation methods and analyze some of the participating metrics. Keywords: machine translation, evaluation, automatic metrics, annotation

The Groningen Meaning Bank (GMB) project develops a corpus with rich syntactic and semantic annotations. Annotations in GMB are generated semi-automatically and stem from two sources: (i) Initial annotations from a set of standard NLP tools, (ii) Corrections/refinements by human annotators. For example, on the part-of-speech level of annotation there are currently 18,000 of those corrections, so called Bits of Wisdom (BOWs). For applying this information to boost the NLP processing we experiment how to use the BOWs in retraining the part-of-speech tagger and found that it can be improved to correct up to 70% of identified errors within held-out data. Moreover an improved tagger helps to raise the performance of the parser. Preferring sentences with a high rate of verified tags in retraining has proven to be the most reliable way. With a simulated active learning experiment using Query-by-Uncertainty (QBU) and Query-by- Committee (QBC) we proved that selectively sampling sentences for retraining yields better results with less data needed than random selection. In an additional pilot study we found that a standard maximum-entropy part-of-speech tagger can be augmented so that it uses already known tags to enhance its tagging decisions on an entire sequence without retraining a new model first. Powered by...

This doctoral thesis is devoted to linguistic analysis of discourse relations as one of the aspects of discourse coherence. Discourse relations are semantic relations holding between propositions in a discourse (discourse arguments). The aim of the thesis is a complex description of discourse relations in Czech and its application in an annotation scheme in the Prague Dependency Treebank. The thesis is divided into three parts: The first one is focused on the theoretical description of discourse relations and on analysis of adequacy of various methodological concepts in corpus processing. The second part describes in detail the proposed scheme for the annotation of discourse relations and the process of the corpus build- up including the evaluation of consistency of the annotated data. Finally, in the last part of the thesis, we address some problematic issues arisen with the employment the proposed scheme and look for their possible solutions.

This work deals with the possibilities of building and exploitation of a corpus of Czech Learners' French. The first part presents the learner corpora in the context of corpus linguistics on the basis of literature and scientific articles relevant to this subject with a special attention for corpora of French L2. Furthermore, the work focusses on the methodological principles of building a learner corpus (e.g. the specific metadata about the respondents and the material) and the error taxonomy. Next section of the work is devoted to analysis of a sample of learner's written productions with a critical description of the error taxonomy.

This master thesis deals with an architecture Model-View-Controller. It explains the principle, model, usage and benefits of an architecture. Futhure more, it presents the best known web frameworks on Java platform which derive benefit from this architecture. There are frameworks like Apache Struts, Tapestry and Spring. In my thesis, the principles of framworks and the work with them are described. Concerning another part of my thesis, there are simple applications which use this frameworks. The work gives instructions to make these applications. It describes and compares the latest versions of these tools. It deals more with framework Spring, describes its elementary principles and modules with enhanced sight to Web MVC of Spring.

One of the problems of semantic web is automated getting annotated data - web pages. Therefore this work is engaged in manual annotation of web pages and try to simplify this process for users using proposed methods. First part contains analysis of annotated data, users and vocabularies used for annotation. Afterwards are proposed methods which simplify annotation creation for users, the possibility of usage similar annotations or possibility highlight interesting parts of web page suitable for annotation. The work includes proposal of annotation tool user interface also that verifies proposed methods in practice. On the basis of this proposal was created a prototype implementation of the annotation tool Semantic Annotator as an extension for Google Chrome browser, which was also used for experiment verifying user-friendliness especially.

Duplicate code occurs in source files for different reasons. In many cases the motivation for copying the code is laziness of a programmer, or an attempt to use an alien source code. Over the years, multiple methods for detection of the duplicate source code have been developed. Approaches vary in the ways they analyze the code, focusing on different representations of the program. Methods based on the analysis of the syntactic properties of the source code often use abstract syntax trees. By examining the tree representation instead of the textual representation of the code, these methods are able to detect duplicate code that underwent formatting changes as well as changes to the names of identifiers. Duplicate code fragments are discovered by identifying the subtrees of the same shape. After the suspicious parts of the tree are identified, further examination of AST properties determines to what extent the code was copied. In this work we develop a system for duplicate code detection based on AST comparison.