An algorithm for generating phrase boundaries for the automatic assignment of prosodic contours in a text-to-speech system for Spanish

Abstract

Determining phonological and intonational phrase boundaries is an important step in synthesizing natural-sounding prosodic contours in a text-to-speech (TTS) system. An algorithm is presented here that generates phrase boundaries in Spanish texts. The basis for this algorithm is Liberman and Church’s function group (f-group) parser for English [M. Y. Liberman and K. W. Church, ‘‘Text Analysis and Word Pronunciation in Text-to-Speech Synthesis,’’ in Advances in Speech Signal Processing, edited by S. Furui and M. M. Sondhi (Marcel Dekker, New York, 1992), pp. 791–831]. The Spanish f-group parser produced satisfactory results for shorter sentences and sentences containing a fair amount of punctuation, but the parser tended to overgenerate phrase boundaries in longer sentences and sentences with little punctuation. In addition, the parser generated but did not distinguish different types of sentence and paragraph transitions. To refine the parser in these areas, selections from a Spanish text [E. Sabato, El túnel (Editorial Seix Barral, Barcelona, 1982)] were read by adult native Spanish speakers and analyzed for phrase boundaries as evidenced by significant changes in fundamental frequency, tempo, and/or pauses. The acoustic phonetic data and its role in the development of the phrase boundary algorithm are discussed.

Design and evaluation of a phonological phrase parser for Spanish text-to-speech

“…The function word lexicon (136 entries) consists of the closed classes of determiners, prepositions, conjunctions, disjunctions; and personal, possessive, and interrogative pronouns. The inclusion of conjunctions and disjunctions in the Function Word lexicon differs from a previous approach 5 , which uses these items to generate major and minor phrase boundaries. Only major phrase boundary placement is considered here.…”

“…Example (5) shows the result of this mislabeling. The verb surgen 'soar' is not in the verb lexicon, so it was labeled a Content Word and grouped with the surrounding Content Words, cables 'cables' and rígidas 'rigid'.…”

A rule-based phrase parser for real-time text-to-speech synthesis

“…Although it has not been tested against a corpus of phrasings,Liberman and Church (1992) observe anecdotally that their parser tends to overgenerate, producing very short, plentiful phrases Karn's (1993). implementation of their approach for Spanish (using her own part of speech tagger) avoids some of these problems by incorporating rules for identifying phrasal conjunctions, relative clauses, and phrase length.…”

Design and evaluation of a phonological phrase parser for Spanish text-to-speech

“…(Both stressed and stressless syllables are included in the syllable count.) This approach differs from a previous approach 5 , which counts syllables in the entire sentence, rather than in the preliminary phrases. To obtain a syllable count, the graphemic text is first converted to phonemes and then syllabified.…”