Data-Driven Pronunciation Generation for ASR Maria Obedkova In ASR systems, dictionaries are usually used to describe pronunciations of words in a language. These dictionaries are typically hand-crafted by linguists. One of the most significant drawbacks of dictionaries created this way is that linguistically motivated pronunciations are not necessarily the optimal ones for ASR. The goal of this research was to explore approaches of data-driven pro- nunciation generation for ASR. We investigated several approaches of lexicon generation and implemented the completely new data-driven solution based on the pronunciation clustering. We proposed an approach for feature extraction and researched different unsupervised methods for pronunciation clustering. We evaluated the proposed approach and compared it with the current hand-crafted dictionary. The proposed data-driven approach could beat the established base- lines but underperformed in comparison to the hand-crafted dictionary which could be due to unsatisfactory features extracted from data or insufficient fine tuning. 1

Although spoken dialogue systems have greatly improved, they still cannot handle communications involving unknown topics. One of the problems is, that they experience difficulties when they should pronounce unknown words. We will investigate methods that can improve spoken dialogue systems by correcting the pronunciation of unknown words. This is a crucial step to provide a better user experience, since for example mispronounced proper nouns are highly undesirable. Incorrect pronunciation is caused by imperfect phonetic representation of the word. We aim to detect incorrectly pronounced words, use knowledge about the pronunciation and user's feedback and correct the transcriptions accordingly. Furthermore, the learned phonetic transcriptions can be added to the speech recognition module's vocabulary. Thus extracting correct pronunciations benefits both speech recognition and text-to-speech components of the dialogue systems.

Le développement d'un système de reconnaissance de la parole exige la disponibilité d'une grande quantité de ressources à savoir, grands corpus de texte et de parole, un dictionnaire de prononciation. Néanmoins, ces ressources ne sont pas disponibles directement pour des dialectes arabes. De ce fait, le développement d'un SRAP pour les dialectes arabes se heurte à de multiples difficultés à savoir, l’'abence de grandes quantités de ressources et l'absence d’'une orthographe standard vu que ces dialectes sont parlés et non écrit. Dans cette perspective, les travaux de cette thèse s’intègrent dans le cadre du développement d’un SRAP pour le dialecte tunisien. Une première partie des contributions consiste à développer une variante de CODA (Conventional Orthography for Arabic Dialectal) pour le dialecte tunisien. En fait, cette convention est conçue dans le but de fournir une description détaillée des directives appliquées au dialecte tunisien. Compte tenu des lignes directives de CODA, nous avons constitué notre corpus nommé TARIC : Corpus de l’interaction des chemins de fer de l’arabe tunisien dans le domaine de la SNCFT. Outre ces ressources, le dictionnaire de prononciation s’impose d’une manière indispensable pour le développement d’un SRAP. À ce propos, dans la deuxième partie des contributions, nous visons la création d’un système nommé conversion (Graphème-Phonème) G2P qui permet de générer automatiquement ce dictionnaire phonétique. Toutes ces ressources décrites avant sont utilisées pour adapter un SRAP pour le MSA du laboratoire LIUM au dialecte tunisien dans le domaine de la SNCFT. L’évaluation de notre système donné lieu WER de 22,6% sur l’ensemble de test. / The development of a speech recognition system requires the availability of a large amount of resources namely, large corpora of text and speech, a dictionary of pronunciation. Nevertheless, these resources are not available directly for Arabic dialects. As a result, the development of a SRAP for Arabic dialects is fraught with many difficulties, namely the lack of large amounts of resources and the absence of a standard spelling as these dialects are spoken and not written. In this perspective, the work of this thesis is part of the development of a SRAP for the Tunisian dialect. A first part of the contributions consists in developing a variant of CODA (Conventional Orthography for Arabic Dialectal) for the Tunisian dialect. In fact, this convention is designed to provide a detailed description of the guidelines applied to the Tunisian dialect. Given the guidelines of CODA, we have created our corpus TARIC: Corpus of the interaction of the railways of the Tunisian Arab in the field of SNCFT. In addition to these resources, the pronunciation dictionary is indispensable for the development of a peech recognition system. In this regard, in the second part of the contributions, we aim at the creation of a system called conversion(Grapheme-Phonème) G2P which allows to automatically generate this phonetic dictionary. All these resources described before are used to adapt a SRAP for the MSA of the LIUM laboratory to the Tunisian dialect in the field of SNCFT. The evaluation of our system gave rise to WER of 22.6% on the test set.