Letter-to-sound rules, also known as grapheme-to-phoneme rules, are important computational tools and have been used for a variety of purposes including word or name lookups for database searches and speech synthesis. These rules are especially useful when integrated into database searches on names and ad-dresses, since they can complement orthographic search algorithms that make use of permutation, deletion, and insertion by allowing for a comparison with the phonetic equivalent. In databases, phonetics can help retrieve a word or a proper name without the user needing to know the correct spelling. A phonetic index is built with the vocabulary of the application. This could be an entire dictionary, or a list of proper names. The searched word is then converted into phonetics and retrieved with its information, if the word is in the phonetic index. This phonetic lookup can be used to retrieve a misspelled word in a dictionary or a database, or in a text editor to suggest corrections. Such rules are also necessary to formalize grapheme-phoneme correspondences in speech synthesis architecture. In text-to-speech systems, these rules are typically used to create phonemes

The automatic derivation of word pronunciations from input text is a central task for any text-to-speech system. For general English text at least, this is often thought to be a solved problem, with manually-derived linguistic rules assumed capable of handling `novel' words missing from the system dictionary. Data-driven methods, based on machine learning of the regularities implicit in a large pronouncing dictionary, have received considerable attention recently but are generally thought to perform less well. However, these tentative beliefs are at best uncertain without powerful methods for comparing text-to-phoneme subsystems. This paper contributes to the development of such methods by comparing the performance of four representative approaches to automatic phonemisation on the same test dictionary. As well as rule-based approaches, three data-driven techniques are evaluated: pronunciation by analogy (PbA), NETspeak and IB1-IG (a modified k-nearest neighbour method). Issues involved in comparative evaluation are detailed and elucidated. The data-driven techniques outperform rules in accuracy of letter-to-phoneme translation by a very significant margin but require aligned text-phoneme training data and are slower. Best translation results are obtained with PbA at approximately 72% words correct on a reasonably large pronouncing dictionary, compared to something like 26% words correct for the rules, indicating that automatic pronunciation of text is not a solved problem.

In this paper, we describe a reversible letter-to-sound/sound-to-letter generation system based on an approach which com-bines a rule-based formalism with data-driven techniques. We adopt a probabilistic parsing strategy to provide a hierarchical lexical analysis of a word, including information such as mor-phology, stress, syllabification, phonemics and graphemics. Long-distance constraints are propagated by enforcing local constraints throughout the hierarchy. Our training and test-ing corpora are derived from the high-frequency portion of the Brown Corpus (10,000 words), augmented with markers indicating stress and word morphology. We evaluated our performance based on an unseen test set. The percentage of nonparsable words for letter-to-sound and sound-to-letter generation were 6 % and 5 % respectively. Of the remaining words our system achieved a word accuracy of 71.8~0 and a phoneme accuracy of 92.5 % for letter-to-sound generation, and a word accuracy of 55.8 % and letter accuracy of 89.4% for sound-to-letter generation. We also compared our hierar-chical approach with an alternative, single-layer approach to demonstrate how the hierarchy provides a parsimonious de-scription for English orthographic-phonological regularities, while simultaneously attaining competitive generation accu-racy.

Dictionary look-up is the primary strategy for deriving pronunciations for input words in a text-to-speech (TTS) system. This strategy is accurate for dictionary words, but it is not complete: it is impossible to list exhaustively all input words. The proper treatment of `unknown' words is currently an unsolved problem in TTS synthesis. There are many competing techniques for letter-to-sound conversion and the system developer must make a rational selection among them. However, it is unclear how different techniques should be properly compared. In this paper, we report a comparative assessment of the competitor methods of letter-to-sound rules, pronunciation by analogy, feedforward neural networks and a k-nearest neighbour method, with respect to their success at automatic phonemisation. This is achieved by using standardised scoring methods, test lexicon and phoneme inventories. The problem of standardising the phoneme set (`harmonisation') is deceptive: this is much harder than at fi...

iii ACKNOWLEDGMENTS I am very pleased to have had the encouragement and support of a committee of three linguists for whom I have the greatest respect and admiration: Mark Liberman, William Labov and Eugene Buckley. Each of them made my transition back to Penn pleasant after what seemed like a long absence. It was a great pleasure to have Mark Randolph both as an external reader and as a colleague at Motorola. Mark’s work at MIT a decade ago has served as an inspiration to me. Orhan Karaali made this dissertation possible in this millennium. As my manager for over two years at Motorola, Orhan insisted on making my dissertation a priority at work. Harry Bliss provided his voice to this project and our whole group is very grateful for his patience and cooperation. My colleagues at Motorola listened to my ideas and provided technical and theoretical assistance at every turn: Noel

this paper are those of Elovitz et al. [2] obtained by anonymous ftp from directory comp.speech/synthesis at svr-ftp.eng.cam.ac.uk. 2.2 Pronunciation by Analogy Pronunciation by analogy (PbA) exploits the phonological knowledge implicitly contained in a dictionary of words and their corresponding pronunciations. The underlying idea is that a pronunciation for an unknown word is assembled by matching substrings of the input to substrings of known, lexical words, hypothesising a partial pronunciation for each matched substring from the phonological knowledge, and concatenating the partial pronunciations. The variant of PbA evaluated here is based on Dedina and Nusbaum's PRONOUNCE [4], but with several further enhancements as detailed by Marchand and Damper [5]. In PRONOUNCE, a data structure called the pronunciation lattice is built from matching substrings in the input word and the dictionary entries. This is a graph containing information about the position and total number of matched substrings, and their partial pronunciations. A possible pronunciation for the input string then corresponds to a complete path through its lattice, with the output string assembled by concatenating the phoneme labels on the nodes/arcs in the order that they are traversed. (Different paths can, of course, correspond to the same pronunciation.) Scoring of candidate pronunciation uses two heuristics in PRONOUNCE. If there is a unique shortest path, then the pronunciation corresponding to this path is taken as the output. If there are tied shortest paths, then the pronunciation corresponding to the best scoring of these is taken as the output. The version of PbA evaluated here features several enhancements over PRONOUNCE. First, we use `full' pattern matching between input letter string and d...

In this thesis, a diphone--based text--to--speech system for Scottish Gaelic, a language spoken by about 80.000 native speakers in Scotland and Canada, is presented. Text-- to--speech systems convert orthographic text input into speech output. The present system consists of two main parts: ffl an automatic phonetic transcription module which produces an orthophonic transcription of the orthographic input text ffl a speech synthesis module which synthesizes an utterance from its transcription by concatenating and modifying previously recorded speech units. Diphones, speech units that cover two sounds and the transition between them, form the basis of the synthesis module. Duration and intonation are modelled on the basis of simple heuristics. The diphone inventory was designed for the Gaelic of Bayble, Lewis. Scottish Gaelic distinguishes four main phonetic settings: velarised, palatalised, nasalised, and neutral. As the domain of these settings is the syllable, they are difficult t...

This paper describes a modification to, and a fast implementation of, the Viterbi algorithm for use in stochastic letter-to-phoneme conversion. A straightforward (but unrealistic) implementation of the Viterbi algorithm has a linear time complexity with respect to the length of the letter string, but quadratic complexity if we additionally consider the number of letter-tophoneme correspondences to be a variable determining the problem size. Since the number of correspondences can be large, processing time is long. If the correspondences are precompiled to a deterministic finite-state automaton to simplify the process of matching to determine state survivors, execution time is reduced by a large multiplicative factor. Speedup is inferred indirectly since the straightforward implementation of Viterbi decoding is too slow for practical comparison, and ranges between about 200 and 4000 depending upon the number of letters processed and the particular correspondences employed in the transdu...

Automatic transliteration problem is to transcribe foreign words in one's own alphabet. Machine generated transliteration can be useful in various applications such as indexing in an information retrieval system and pronunciation synthesis in a text-to-speech system. In this paper we present a model for statistical Englishto -Korean transliteration that generates transliteration candidates with probability. The model is designed to utilize various information sources by extending a conventional Markov window. Also, an efficient and accurate method for alignment and syllabification of pronunciation units is described. The experimental results show a recall of 0.939 for trained words and 0.875 for untrained words when the best 10 candidates are considered.