|
|
Assessing the impact of manual corrections in the Groningen Meaning Bank
Weck, Benno ; Lopatková, Markéta (advisor) ; Vidová Hladká, Barbora (referee)
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...
|
|
|
Spanish similes with adjective and it's equivalents in Czech
Šilhavá, Terezie ; Čermák, Petr (advisor) ; Kratochvílová, Dana (referee)
This dissertation focuses on Spanish similes with adjective (like feliz como una perdiz) and its equivalents in Czech. First the theoretical frame is sketched: a summary of the evolution and of the actual situation of the investigation in the field of fraseology; a description of the corpora which were used as a source of the data; a characterization of the basic terms connected with the topic; a description of the function of phrasemes and simies in the process of communication; and a summary of some basic thoughts from cognitive linguistics. Then a semantic and contrastive analysis of the simies with selected adjectives (feliz, alegre, triste, tranquilo, furioso) comes. Finally we point out what each comparatum can tell about the meaning of the adjective, what we can tell about the comparata and its selection, and what can be said about the grade of stabilization of the examined simies. Then a tip follows: what more could be explored in this area and that there could be a possibility to use the results of such research for instance while creating a didactic material.
|
|
| |
|
| |
|
|
The presentation of speech, writing and thought in British newspapers.
Čermáková, Barbora ; Malá, Markéta (advisor) ; Dušková, Libuše (referee)
The thesis compares forms of speech presentation (and marginally writing presentation) in British newspapers through a theory of Speech, Writing and Thought presentation as introduced by E. Semino and M. Short (2004). On a specialised corpus of 6 newspaper articles, the work compares qualitive and quantitative tendencies of individual speech presentation categories, focusing on the diagnosis of differences and common points in two journalistic sub-genres, the tabloid and the broadsheet. The speech presentation categorization is applied as a clinal model with permeable borders between individual categories. The focus is on detecting and analysing the fucntions of individual categories of speech presentation in terms of information, style, pragmatics and form. Powered by TCPDF (www.tcpdf.org)
|
|
|
Multiword expressions in Italian
Jungwirthová, Klára ; Štichauer, Pavel (advisor) ; Obstová, Zora (referee)
The main topic of this thesis are the multiword expressions in the italian language. The thesis is divided into two parts - the theorical and the empirical part. The theorical part deals with the multiword expressions, the syntagmas and the idiomatic expressions. In the empirical part the connections between the constituents of the multiword expressions will be researched. Than four criteria will be on the multiword expressions applied (head inflection, insertion of the head's modifiers, pronominalisation of the head and dislocation and topicalization of the head). These transformations will be verified with the aid of corpora and questionnaires. Depending on the results of this research will be decided if the multiword expressions resemble the syntagmas or the idiomatic expressions.
|
|
|
Intertextuality in Advertising
Žbirková, Nikola ; Mareš, Petr (advisor) ; Šebesta, Karel (referee)
This bachelor thesis examines the intertextuality in advertising. The theoretical part first of all defines the basic concepts, such as advertising, intertextuality and its types or transtextuality, and in addition includes selected approaches to intertextuality in advertising and their subsequent comparison. The practical part contains a design of an information system for working with a corpus and the introduction of the corpus itself, accompanied by comments and evaluation. The final section consists of an analysis of the corpus, which leads to the formulation of conclusions about how the source texts are modified and adapted. This thesis aims to: 1) find out how frequent is the advertisement based on the principle of intertextuality, 2) ascertain what means of intertextuality are the most utilized in advertising texts, 3) determine what type of source text is the most frequently referred, 4) design an appropriate information system to store, search easily and comment on the collected material.
|
|
|
The adverbs "absolutely", "completely" and "totally" in contemporary American soap operas
Gižová, Veronika ; Malá, Markéta (advisor) ; Popelíková, Jiřina (referee)
This thesis is concerned with the functions of adverbs absolutely, completely and totally. These adverbs are generally employed as adverbials denoting degree or intensifiers with gradable adjectives and adverbs. Their use is most frequent in informal spoken language due to their ability to occur as an emphatic agreement in the form of ellipsis. The distribution of the selected adverbs differs in inspected functions - the secondary sources claim the adverb totally has been used increasingly as an elliptical agreement. The results furthermore demonstrate their varying collocations and constructions in which they occur. The adverbs are examined on the corpus of contemporary American soap opera Friends whose language closely resembles informal dialogue. A combination of corpus analysis methods are employed in order to determine the status of the syntactic constructions containing absolutely, completely and totally. The analysis of 120 examples has proven that while absolutely and completely both occur primarily in one function, different for each adverb,, totally may be used in a variety of detected functions resulting in semantic blends.
|
|
|
Error Analysis of texts written by non-native speakers of Czech
Pfeiferová, Jana ; Šormová, Kateřina (advisor) ; Pierścieniak, Piotr Paweł (referee)
The bachelor thesis deals with the use of language of non-native speakers of the Czech language - specifically students whose first language is Russian. The first part focuses on different approaches to language errors in terms of the theories which deal with the acquisition of the second language. In the second part, written texts of non-native speakers of the Czech language are analyzed by the method of error analysis. Errors are separately analyzed from a grammatical, morphologic, syntactic and lexicological perspective. Then errors are divided into categories and consequently described and interpreted. Interpretation of the errors takes into account the frequency of errors and focuses on intralingual errors and interlingual errors.
|
|
|
Spanish determiners este, ese, aquel and their Czech equivalents
Erbenová, Adéla ; Čermák, Petr (advisor) ; Kratochvílová, Dana (referee)
The thesis is focused on the Spanish determiners este, ese, aquel, the definite article and their anaphoric function. First, the theoretical groundwork for the latter analysis is built: the deter- miners are characterized together with the anaphora as a device of discourse coherence. Then, the choice between the demonstratives este, ese, aquel and the defi- nite article in different types of nominal anaphora is analyzed using examples from the parallel corpus InterCorp. Diverse factors influencing this choice are examined. The analysis is extended with a comparative analysis in which Czech equivalents of the Spanish determiners in the same types of anaphora are compared.
|
guest ::
