Notre thèse s’intéresse aux processus de création des adverbes de modalité dubitative dans l’histoire de l’espagnol. À l’aide des corpus de la Real Academia Española, CORDE et CREA, nous étudions du point de vue diachronique les principaux adverbes et locutions adverbiales qui expriment le doute du locuteur par rapport au contenu propositionnel de l’énoncé : por (a)ventura, acaso, quizá(s), puede que, tal vez, a lo mejor, lo mismo et igual. Notre recherche comprend ainsi la quasi-totalité de l’histoire de la langue espagnole, depuis le XIIe siècle, où l’on retrouve déjà quizá, jusqu’au XXe avec le surgissement de lo mismo et igual. Ces expressions ont des origines diverses – constructions verbales, syntagmes prépositionnels, etc.–mais elles fonctionnent toutes initialement comme des éléments intégrés dans la structure syntaxique de la phrase, dans laquelle elles transmettent des notions qui relèvent du hasard, de la temporalité, de la comparaison, etc. Ce travail s’occupe donc, dans un premier temps, d’élucider les chemins que ces expressions ont empruntés pour parvenir à fonctionner comme des adverbes de modalité dubitative et, dans un deuxième temps, il rend compte des relations qu’entretiennent ces expressions adverbiales, puisque, malgré leur apparente synonymie et interchangeabilité, ces adverbes diffèrent sur certains aspects les uns des autres. À cet égard, l’étude diachronique des contextes d’apparition, des valeurs et des fonctions des éléments qui ont intégré le paradigme nous permet de comprendre les différentes nuances et possibilités d’emploi qui caractérisent ces expressions une fois qu’elles acquièrent leur rôle d’adverbes de modalité dubitative. / This thesis aims to study the creation processes of the modal probability adverbs in the history of Spanish. Thus, from the diachronic point of view, we study, with the help of the Real Academia Española corpora, CORDE and CREA, the main adverbs and adverbial locutions that express the speaker's doubt regarding the propositional content of the statement: por (a)ventura, acaso, quizá(s), puede que, tal vez, a lo mejor, lo mismo and igual. Our research covers practically the whole history of the Spanish language, from the twelfth century, in which quizá is documented, until the twentieth, when the most recent adverbs of probability, lo mismo and igual, arise. These expressions have different origins – verbal constructions, prepositional phrases, etc. – but all function initially as integrated elements in the syntactic structure of the sentence, in which they transmit notions related to chance, temporality, comparison, etc. This work, therefore, is concerned firstly with elucidating the paths followed by these expressions in order to function as probability adverbs and, secondly, it gives account of the relations that lie between these adverbial expressions. Despite their apparent synonymy and interchangeability, these adverbs differ from one another in some respects. In this sense, the diachronic study of the contexts of appearance, the meanings and the functions of the elements that have integrated the paradigm allow us to understand the different nuances and possibilities of use that characterize these expressions once they acquire their role of modal probability adverbs. / Esta tesis se ocupa del estudio de los procesos de creación de los adverbios de modalidad dubitativa en la historia del español. Con la ayuda de los corpus de la Real Academia Española, CORDE y CREA, estudiamos desde el punto de vista diacrónico los principales adverbios y locuciones adverbiales que expresan la duda del hablante con respecto al contenido proposicional del enunciado: por (a)ventura, acaso, quizá(s), puede que, tal vez, a lo mejor, lo mismo e igual. Nuestra investigación abarca así prácticamente toda la historia de la lengua española, desde el siglo XII, en el que ya se documenta quizá, hasta el XX, en el que surgen los adverbios de duda más recientes lo mismo e igual. Las expresiones estudiadas tienen orígenes diversos –construcciones verbales, sintagmas preposicionales, etc.–, pero todas funcionan inicialmente como elementos integrados en la estructura sintáctica de la frase, en la que transmiten nociones relativas al azar, a la temporalidad, a la comparación, etc. Este trabajo se ocupa, pues, en primer lugar, de elucidar los caminos que han seguido estas expresiones para llegar a funcionar como adverbios de modalidad dubitativa y, en segundo lugar, de dar cuenta de las relaciones que mantienen estas expresiones adverbiales, ya que, a pesar de su aparente sinonimia e intercambiabilidad, estos adverbios difieren unos de otros en algunos aspectos. En este sentido, el estudio diacrónico de los contextos de aparición, valores y funciones de los elementos que han integrado el paradigma nos permite comprender los diferentes matices y posibilidades de uso que caracterizan a estas expresiones una vez que adquieren su papel de adverbios de modalidad dubitativa.

Several studies point out different causes of performance degradation in supervised machine learning. Problems such as class imbalance, overlapping, small-disjuncts, noisy labels, and sparseness limit accuracy in classification algorithms. Even though a number of approaches either in the form of a methodology or an algorithm try to minimize performance degradation, they have been isolated efforts with limited scope. This research consists of three main parts: In the first part, a novel probabilistic diagnostic model based on identifying signs and symptoms of each problem is presented. Secondly, the behavior and performance of several supervised algorithms are studied when training sets have such problems. Therefore, prediction of success for treatments can be estimated across classifiers. Finally, a probabilistic sampling technique based on training set diagnosis for avoiding classifier degradation is proposed<br>

本論文依據現代客語語料庫，探討臺灣客語評注性情態副詞在語法化、詞彙化以及語用化運作之下的發展過程，以及情態與語言演變機制之間的關係。藉由三個知識性情態詞，本文對於評注性情態副詞之分佈、語義和句法特徵進行詳細的描述。藉由檢視其構式結構以及運用相關測試，證實評注性情態副詞不但遵守副詞修飾範域：話語行為副詞 > 情態副詞 (評論副詞 > 知識性情態副詞) > 附加副詞，也同樣依循情態詞之階層分佈：知識性情態詞 > 義務性情態詞 > 動力性情態詞。在他們所處之語意語境以及語用語境影響之下，語用推論的產生導引出重新分析，此轉喻強化現象，致使評注性情態副詞在階段發展中經歷了語法及語意之轉變。在語言使用之溝通目的下，評注性情態副詞透過語境誘發之重新解釋形成了歧義現象，並進一步固定化為獨立之語意。綜言之，奠基於臺灣客語語料庫之實際語料，本文提供了對評注性情態副詞全面的深入探索。研究結果具有兩項主要貢獻：將語法化，詞彙化和語用化的理論框架納入臺灣客語情態副詞之分析；並藉由語言實證展示情態與主觀性之間相互關係之理論模式。 / This dissertation takes a corpus-based approach in an attempt to explain the development of Taiwan Hakka modality disjuncts with epistemic interpretation on the theoretical ground of the mechanisms of linguistic change, including grammaticalization, lexicalization, and pragmaticalization. The main focus of this research is on the interaction between modality and the three mechanisms. A fine-grained analysis is provided to examine the morphological characteristics and syntactic structures of epistemic and evaluative modality disjuncts, exemplified by three cases in Taiwan Hakka. Several constructions and tests are manipulated to figure out the scope of modal disjuncts, which warrant the postulation of the scope hierarchy of illocutionary disjunct > modality disjunct (evaluative > epistemic) > adjunct, as well as the ranking of epistemic modal > deontic modal > dynamic modal. Thanks to their semantic-pragmatic contexts, the operation of pragmatic inferences lead them into reanalysis, giving rise to grammatical and semantic shift from one stage to the next. Metonymic strengthening is claimed to be the key to motivate such developments. The usage in a context triggers context-induced reinterpretation, resulting in ambiguity. Overall, this dissertation, probing into actual language usages from corpora, proposes a full-fledged analysis on modality disjuncts. The findings feature two significant contributions—an integration of the theoretical frameworks of grammaticalization, lexicalization and pragmaticalization into the analysis of Taiwan Hakka modality disjuncts; and a demonstration of linguistic investigation for the theorization on the relationship between modality and subjectivity.