This study investigates the application of machine learning (ML) techniques in detecting fake news, addressing the rapid spread of misinformation across social media platforms. Given the time-consuming nature of manual fact-checking, this research compares the robustness of basic machine learning models, such as Multinominal Naive Bayes classifiers, with complex models like Distil-BERT in identifying fake news. Utilizing datasets including LIAR, ISOT, and GM, this study will evaluate these models based on standard classification metrics both in single domain and cross-domain scenarios, especially when processing linguistically manipulated data. Results indicate that while complex models like Distil-BERT perform better in single-domain classifications, the Baseline models show competitive performance in cross-domain and on the manipulated dataset. However both models struggle with the manipulated dataset, highlighting a critical area for improvement in fake news detection algorithms and methods. In conclusion, the findings suggest that while both basic and complex models have their strength in certain settings, significant advancements are needed to improve against linguistic manipulations, ensuring reliable detection of fake news across varied contexts before consideration of public availability of automated classification.

Cette recherche présente une méthode d'analyse micro-systémique des mots composés thaïs. Le but denotre étude est de trouver une réponse au questionnement suivant « existe-t- il une voie qui permette de traduireautomatiquement les mots thaïs vers le français avec un résultat parfait ? ». Ce travail est divisé en cinqchapitres. La première partie concerne une histoire brève de la traduction automatique dont celle du thaï. Lespoints de vue des autres travaux sont étudiés. Le deuxième chapitre présente les caractéristiques de la langue thaïe qui possède une forme d’écriture typique sans espacement et peut entrainer des difficultés en termes d’ambiguïté dans la traduction. Certaines divergences entre le thaï et le français sont soulignées à l’aide de la théorie micro-systémique du Centre Tesnière. Le troisième chapitre fait l’étude des mots composés thaïs en utilisant une méthode hybride de l’analyse morphosyntaxique et notre système à base de règles conformes à notre modèle d'analyse de données. Le quatrième chapitre met en évidence un contrôle modélisé des unités lexicales codées syntaxiquement et sémantiquement afin d’en définir des algorithmes efficaces. Le dernier chapitre conclut sur les résultats des nouveaux algorithmes par leur informatisation. Sont enfin énoncées les perspectives ouvertes par cette nouvelle recherche. Cette étude est présentée comme un travail fiable à l’élimination des ambiguïtés. Fondée sur une méthode hybride, elle nous a permis d’atteindre notre objectif et de trouver ainsi une voie efficace qui nous autorise à traduire automatiquement les mots thaïs vers le français. Le résultat place cet outil comme l’un des plus accessibles à la recherche internationale où le thaï et le français prennent leurs places de choix / This thesis presents a method of Micro-Systemic Linguistic Analysis of Thai compound words. The aim of our study is to find out: “Is there any method which allows us to translate Thai words into French automatically with a perfect result?” Our work is divided into five chapters as follows : The first chapter concerns a brief a history of machine translation including that of the Thai. Some notable points of view are discussed. The second chapter identifies some essential characteristics of the Thai language such as a non-space writing style resulted in ambiguity in machine translation. Different entities between Thai and French languages are underlined by means of the micro-systematic theory of the Centre Tesnière. The third chapter analyses Thai compound words using a hybrid method involving morpho-syntactic parsing and a rule-based system corresponding to our model of data analysis. The fourth chapter employs a technique of lexical-syntactic and semantic control enabling the definition of efficient algorithms. The final chapter concludes our work with some future perspectives. This study is presented as a reliable approach which enhances the elimination of word ambiguities in machine translation. This hybrid method allows us to reach our objective and to find an effective way to translate Thai to French automatically. The result could be an accessible tool for international research in the Thai and French languages

L’objet de la présente recherche vise à donner un éclairage théorique, méthodologique et pratique de l’augmentation de capital étudiée dans un contexte de crise. En effet, le cadre conceptuel de cette recherche a porté sur les termes suivants : augmentation de capital, difficulté financière, signaux annonciateurs de la situation difficile, notion d’entreprise en situation difficile et plus particulièrement sur la théorie de signalisation et sur la théorie de la lecture optionnelle du bilan. Il convient d’affirmer que ce cadre théorique a eu pour intérêt de « contextualiser » et de délimiter le sujet de recherche. Pour affiner la délimitation du sujet, une étude exploratoire a été réalisée. Elle a permis d’identifier les indicateurs de la situation difficile et de tester la méthodologie d’étude d’événement sur un cas. Enfin, pour valider cette recherche et pour permettre la généralisation des résultats, la méthodologie d’étude d’événement a été appliquée à un échantillon de 30 cas d’entreprises en situation difficile. Le but de cette étude a été de mesurer la création ou la destruction de la valeur actionnariale observable lors du lancement de ce type d’opération qui permet de mesurer les réactions des actionnaires anciens. Par ailleurs, pour mieux appréhender la perception des investisseurs sur ce type d’opération, la recherche s’est orientée vers les analyses de sens et textuelles lexicales. Au final, le sujet étudié peut s’appréhender comme un financement de dernier recours pour les sociétés en situation difficile. / The aim of the present research is to provide a theoretical lighting, methodology and practical of the capital increase studied in a context of crisis. Indeed, the conceptual framework of this research focused on the precision of terms such as capital increase, financial difficulty, covenant of difficulty, concept of enterprise in difficulty and in particular on the indication of theories namely signaling and the balance sheet optional reading. We wish to emphasize that this framework has had interest to « contextualize » and to delimit the research topic. And to improve the delimitation of the topic, an exploratory study was conducted and it has identified the indicators of difficulty and test the event study methodology on a case. Finally, to validate this research and to allow generalization of results, event study methodology was applied to a sample of 30 cases of companies in difficulty. The aim of study was to measure the creation or destruction of shareholder value observable at the launch of this type of operation to measure the reactions of former shareholders. Furthermore, to better understand the perception of investors in this type of operation, the research is oriented analyzes of lexical meaning and textual. In the end, the topic studied can be apprehended as a last resort financing for companies in difficulty.

p, li { white-space: pre-wrap; } The main goal of this work is to analyze the creation of the composite compiler. Composite compiler is in this case a szstem, which consists of more cooperating parts. My compiler is special, because its syntactic analyser consists of two parts. The work is focused on the construction of the parsers parts, on its cooperation and comunication. I will trys to scatch the teoretical backgroun of this solution. This is to be done by gramatical systems. Then I~will try to justify whether or not it is neccesary and suitable to create such a kind of parser. Last but not least I~will analyse the language, whose syntactic analyser is to be implemented by the chosen method.

The subject of the present thesis is a penitential Nekotoraja zapoved which is a collection of sins and penance designated for confessors. The writing is preserved in a Russian redaction of Church Slavonic from 14th-15th century, but the Czech origin of 11th century was already proved by philological analysis. The thesis presents Nekotoraja zapoved with all its specifics and in the context of the Church Slavic literature. A miscellaneous content of the penitential reflects cultural, historical, social and political aspects of the medieval society. The aim of this thesis is a philological edition, a lexical analysis of almost hundred lexemes which characterizes the lexicon of Nekotoraja zapoved and a comparison with Great Moravian penitential Zapovedi svętychЪ otc. The comparison of both contents shows changes of confessional convention by then and the lexical comparison reflects possibilities of the Old Slavonic lexemes usage and variations of expressions.

This master's thesis deals with implementation of ISO C99 language interpreter. The goal of this thesis is to provide support of education in C language programming and fast algorithm prototyping. It enables students to create own C programs and to experiment with language constructions without compiling. User interface includes editor and simple debugger. The interpreter is implemented in a novel grammar development environment written in Java language - ANTLRWorks which includes ANTLR language tool.

The Translation Technique from Latin to Old Church Slavonic on the Basis of Gumpold's Legend and the Second Old Church Slavonic Life of Saint Wenceslas Kateřina Spurná Summary This dissertation analyses the translation technique from Latin to Old Church Slavonic on the basis of Gumpold's Legend (Gump) from the end of the 10th century and its translation, the Second Old Church Slavonic Life of Saint Wenceslas (VencNik), which was written in early Přemyslid Bohemia probably in the second half of the 11th century. The introductory chapters provide the basic characteristics of the Gump and deal with its preservation in manuscripts and other Latin and Old Church Slavonic legends of St Wenceslas. Afterwards, attention is focused on the analysis of the VencNik, its preservation in manuscripts of Russian provenance and the basic phonetic and morphological characteristics of the text. These chapters are followed by a new edition of the VencNik, in which the Old Church Slavonic text is presented in the form actually preserved in manuscripts (unlike the earlier edition of Josef Vašica from 1929, who tried to reconstruct the original version), and compared with the partly revised and supplemented edition of the Gump. The edition of the Old Church Slavonic text is followed by a brief chapter on the biblical quotations...