RealSpeak Telecom Software Development Kit

Transcription

1 RealSpeak Telecom Software Development Kit User s Guide for Castilian Spanish V4.0

2 EV AL U AT I O N AG R E E M E N T READ THE FOLLOWING CAREFULLY BEFORE USING THE SOFTWARE. PRIOR TO RECEIVING THE SOFTWARE, YOU HAVE SIGNED EITHER A LICENSE AGREEMENT OR A NON-DISCLOSURE AGREEMENT WITH SCANSOFT, CONCERNING THE REALSPEAK TELECOM DEVELOPMENT SYSTEM OR SOME OF ITS COMPONENTS. THEREFORE, REFERENCE IS MADE TO THESE AND AS SUCH THE CONTRACTUAL TERMS AND CONDITIONS SHALL APPLY TO THE SOFTWARE. IF YOU HAVE NOT SIGNED SUCH AN AGREEMENT AND IF YOU USE THE SOFTWARE, SCANSOFT WILL ASSUME THAT YOU AGREED TO BE BOUND BY THE EVALUATION AGREEMENT SPECIFIED HEREUNDER. IF YOU DO NOT ACCEPT THE TERMS OF THIS EVALUATION AGREEMENT, YOU MUST RETURN THE PACKAGE UNUSED TO SCANSOFT WITHIN SEVEN (7) DAYS AFTER RECEIPT. 1. Grant of Rights In consideration of a possible commercial relationship, ScanSoft hereby grants to you, the LICENSEE, who accepts, a non-exclusive right to internally evaluate and test the software program ( the Software ). 2. Ownership of Software ScanSoft retains title, interests and ownership of the Software recorded on the original disk(s) and all subsequent copies of the Software and Documentation, regardless of the form or media in or on which the original and other copies may exist. ScanSoft reserves all rights not expressly granted to LICENSEE. 3. Copy Restrictions This Software and the accompanying documentation are copyrighted. Unauthorized copying of the Software, including Software that has been merged or included with other software, or of the documentation is expressly forbidden. LICENSEE may be held legally responsible for any intellectual property infringement that is caused or encouraged by his failure to abide by the terms of this agreement. LICENSEE is allowed to make two (2) copies of the Software solely for backup purposes, provided that the copyright notice is included on the backup copy. 4. Use Restrictions LICENSEE agrees not to use the Software for any other purpose than internally evaluating the Software. LICENSEE may physically transfer the Software from one computer to another, provided that the Software is used on only one computer at a time. LICENSEE may not modify, adapt, translate, reverse engineer, decompile, disassemble or create derivative works based on the Software.LICENSEE may not modify, adapt, translate or create derivative works based on the documentation provided by ScanSoft. The Software may not be transferred to anyone without the prior written consent of ScanSoft. In no event may LICENSEE transfer, assign, lease, sell or otherwise dispose of the Software and Documentation on a temporary or permanent basis except as expressly provided herein. 5. Warranty THE SOFTWARE IS PROVIDED AS IS WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING WITHOUT LIMITATION WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. ScanSoft shall have no liability to LICENSEE or any third party for any claim, loss or damage of any kind, including but not limited to lost profits, punitive, incidental, consequential or special damages, arising out of or in connection with the use or performance of the Software and accompanying documentation. 6. Termination This agreement is effective until terminated. ScanSoft reserves the right to terminate this agreement automatically if any provision of this agreement is violated. LICENSEE may terminate this agreement by returning the Software and the accompanying documentation to ScanSoft, along with a written warranty stating that all copies have been returned.

3 Copyright (C) ScanSoft, Inc RealSpeak Telecom Software Development Kit User s Guide for Castilian Spanish V4.0 - December 2004-December 2005 No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording or any information retrieval system, without the written permission of ScanSoft. Trademarks MS-DOS, WINDOWS, MICROSOFT VISUAL C++, BORLAND C++ and Sound Blaster are registered trademarks of their respective owners. ScanSoft is a registered trademark. All rights reserved.

4 Document History Date Type of change SoftwareVersion November 2004 December 2005 Reformatting of document V4.0 Update for version 4.0 Update SSML V4.0 Preprocesor chapter and Using Control Sequences section of Chapter I

5 Table of Contents CASTILIAN SPANISH TEXT-TO-SPEECH SYSTEM... 8 Introduction...8 Preparing a text for Text-To-Speech...8 Native Character Set...8 Using Control Sequences...8 Quick Reference of the RealSpeak native Control Sequences for Castilian Spanish...10 Speech Rates in Words per Minute for Castilian Spanish Voices...14 Entering phonetic input...14 How to proceed...14 Lexical stress and sentence accents in phonetic input...15 The Castilian Spanish L&H+ and UNIPA Phonetic Alphabets...17 Using a User Dictionary...19 Using the Microsoft SAPI5 Lexicon...19 User Lexicons...19 Application Lexicons...19 The Castilian Spanish SAPI5 Phoneme List...20 Notes on the Castilian Spanish Text-To-Speech System...22 Cardinal Numbers...22 Decimal Numbers...22 Ordinal Numbers...23 Roman Numbers...23 Bank Account Numbers...24 Telephone Numbers...24 Dates...25 Time Indications...25 Currencies...26 Abbreviations...26 Acronyms PREPROCESSOR Introduction Header Processing...30 Header Field Extraction...30 Header Field Reading...32 From Field...32 Date Field...33 Subject Field body processing...34 Message Extraction...34 Text Normalization...36 Language Specific Text Normalization...38 English words...38 Customizing the Preprocessor...40 Customizing the Header...40 From Field...40 Subject Field...43 Customizing the Body...44

6 SSML PREPROCESSOR Introduction...46 Castilian Spanish specific SSML markup...46 XML encoding types for Castilian Spanish SML Specifics for Castilian Spanish...47 CUSTOM G2P DICTIONARIES Introduction...52 APPENDICES Appendix A: Castilian Spanish voice names...54

7 Realspeak Telecom SDK Chapter I Castilian Spanish Text-To-Speech System User s Guidefor Castilian Spanish V4.0

8 Chapter I Castilian Spanish Text-to-Speech system Castilian Spanish Text-To-Speech System Introduction This section provides operational instructions for the Text-To-Speech system for Castilian Spanish. It reviews the functionality of the system, and describes how the user can customize the pronunciation of input texts. This part also describes issues that are particular to the Castilian Spanish Text-To-Speech system. It introduces the Castilian Spanish phonetic alphabet and discusses some language-specific features of the Castilian Spanish Text-To-Speech system. Preparing a text for Text-To-Speech Native Character Set Using Control Sequences In general, there are four ways to intervene in the pronunciation of text: By using control sequences By entering phonetic input By using a user dictionary or a user ruleset By using one of the supported API s These mechanisms are described in the Programmer s Guide. In this part, however, the specifications for Castilian Spanish are fully described. The native character set of the Castilian Spanish TTS system is Windows-1252; it has the printable characters in the ASCII range as a subset. Note that TTS input encoded in another supported character set is converted to the native character set for that language before it is processed internally. Consequently, input must be representable in the native character set even if it is encoded in another character set supported by the API. For a description of the various supported markup languages (independent from the language), refer to the Programmer's Guide. Remark: <ESC> represents the escape character \x1b (decimal 27) that generates the ASCII character 27 (Hex 1B). User's Guide for Castilian Spanish Chapter I/8

9 Chapter I Castilian Spanish Text-to-Speech system Below, you find a quick reference table for the RealSpeak native control sequences for Castilian Spanish. The language-specific support for the SSML markup language is described in the SSML Preprocesor chapter. User's Guide for Castilian Spanish Chapter I/9

10 Chapter I Castilian Spanish Text-to-Speech system Quick Reference of the RealSpeak native Control Sequences for Castilian Spanish Sequence Description Range Default Delimiter <ESC> \vol=x\ <ESC> \rate=x\ <ESC> \rate_wpm= xxx\ <ESC>Mx Volume (x : ) 0 = silence 10 = low 100 = high 80 No <ESC>\vol=10\ Puedo hablar con una voz baja, <ESC>\vol=90\ pero también con una voz alta. Speech Rate (x : ) 10 = slow 100 = fast 50 No Puedo hablar <ESC>\rate=70\ con una velocidad más rápida <ESC>\rate=20\ o una velocidad más lenta. Words per minute (xxx: ) Voice-specific (see subsequent table) Voicespecific No Puedo hablar <ESC>\rate_wpm=350\ con una velocidad más rápida <ESC>\rate_wpm=110\ o una velocidad más lenta. Read mode; some read modes are not supported in e- mail mode x = 0..3: 0 = character-bycharacter 1 = word-by-word (not supported in e- mail mode) 2 = sentence-bysentence 3 = line-by-line (not supported in mode) <ESC>M0 Demostrador (The word "Demostrador" will be spelled.) <ESC>M1 Esto es el Demostrador español. (This sentence will be read word by word.) <ESC>M2 Esto es el Demostrador español. (This sentence will be read as one sentence.) 2 Yes User's Guide for Castilian Spanish Chapter I/10

11 Chapter I Castilian Spanish Text-to-Speech system Sequence Description Range Default Delimiter <ESC>Wx Wait Period 0 = no wait period 1 = 200 milliseconds wait period 9 = 1800 milliseconds 2 Yes <ESC>M2 <ESC>W2 Esta frase está seguida de una pausa corta. <ESC>W9 Esta frase está seguida de una pausa larga. Son distintos? <ESC> \Pause=xxx \ Long Pause msec No Puede insertar una pausa entre los mensajes especificando la duración <ESC>\Pause=1500\ de la pausa. <ESC>" Sentence Accent <ESC>"Cristina viene mañana. Cristina viene <ESC>"mañana. (i.e. no Carmen) (i.e. no hoy) No <ESC>C <ESC>E Note: Manually inserted sentence accents may have no effect in RealSpeak. The RealSpeak synthesis module may indeed have reasons to override the requested sentence accent, and thus not realize it. Continuation Yo como manzanas. Y peras también. Yo como manzanas. <ESC>C Y peras también. No In the first of the above examples, the text-to-speech system will detect an end-of-sentence after manzanas and hence pause before pronouncing the second part of the input. In order to make the system pronounce the entire input as one sentence, a continuation sequence should be inserted. End-of-Message Yes Esta es la primera parte de la frase <ESC>E y ésta es la segunda <ESC>/+ In the above example, the sequence <ESC>E forces the system to read the entire input as two separate sentences. Phonetic Input (L&H+ phonetic alphabet) No User's Guide for Castilian Spanish Chapter I/11

12 Chapter I Castilian Spanish Text-to-Speech system Sequence Description Range Default Delimiter <ESC>/+la_'t&Si.ka<ESC>/+ <ESC>%x <ESC>F <ESC>\tn= s\ Preprocessing Mode text = standard text mode = mode Yes <ESC>%text Tiene usted un mensaje enviado por Luís Morales. <ESC>% From: "Luís Morales" <lmorales@arrakis.es> Reset to Default Yes <ESC>\vol=90\ Es el volumen más alto. <ESC>F Es el volumen por defecto. <ESC>\rate=10\ Es la velocidad más baja. <ESC>F Es la velocidad por defecto. Guide text normalization; limited support in mode address=address mode (not supported in mode) normal=standard mode spell=spell mode <ESC>\tn=spell\spell <ESC>\tn=normal\spell Comment: The text normalization types corresponding with the SSML <sayas> types are also supported in standard text mode (not in mode), see the SSML Preprocesor chapter for more details. <ESC>@c <ESC>\domain =s\ <ESC>\domain\ <ESC>\voic e=s\ <ESC>\mrk =n\ Declare the part-of-speech (not supported in Spanish) Enable the extension (only if a custom g2p has been loaded) Disable the last extension Set the voice c is a character s = string: the name of the extension s = string: the name of the voice <ESC>\voice=Isabel\Mi nombre es Isabel. Insert a n = bookmark <ESC>\mrk=1245\Insertarlo. User's Guide for Castilian Spanish Chapter I/12 No Yes Yes Yes No

13 Chapter I Castilian Spanish Text-to-Speech system Sequence Description Range Default Delimiter <ESC>\p\ Insert a paragraph boundary Yes Yo soy <ESC>\p\Isabel. <ESC>\aud Insert an audio s = string: Yes io="s"\ file; not the URI of a supported in e- document with an mail mode appropriate MIME type <ESC>\audio="s0001.wav"\Insertarlo User's Guide for Castilian Spanish Chapter I/13

14 Castilian Spanish Text-to-Speech system Chapter I Speech Rates in Words per Minute for Castilian Spanish Voices Words per minute Voice Range Default Isabel Min = 82 Max = Entering phonetic input How to proceed To switch from orthographic to phonetic mode, insert <ESC>/+ to use the L&H+ phonetic alphabet. The phonetic input mode remains active until the command is explicitly reset by entering <ESC>/+ again. The phonetic input string is composed of symbols of the L&H+ phonetic alphabet (see phonetic table below). Examples are given in the phonetic table below. In addition to the phonetic symbols, it is advised to use the following characters in the phonetic input string: L&H + Symbol ' (ASCII 39, Hex 27) " (ASCII 34, Hex 22) Meaning primary word stress sentence accent Special characters As in: <ESC>/+ tr6a.mi.te <ESC>/+ (noun 'trámite') vs. <ESC>/+ tr6a. mi.te <ESC>/+ (verb form from 'tramitar') <ESC>/+?aj_ Dos_a. Ten.tos_en_la_ fr6a.se <ESC>/+ (Hay dos acentos en la frase.). syllable boundary <ESC>/+ 'si.la.ba <ESC>/+ (sílaba) # silence (pause) <<ESC>/+ di.xo#_no.lo. a.gas <ESC>/+ (Dijo: no lo hagas.) User's Guide for Castilian Spanish Chapter I/14

15 Castilian Spanish Text-to-Speech system Chapter I Note that the use of punctuation marks remains useful within phonetic input to assure a correct intonation. Each punctuation mark needs to be preceded by an asterisk. <ESC>/+?a. Jer6*, kwan.do.ter6.mi. na.r6on*, se. fwe.r6on.a. ka.sa*. <ESC>/+ L&H+ Symbol _ Punctuation Marks Meaning Word delimiter *. End of declarative *, Comma *! End of exclamation *? End of question *; Semicolon *: Colon Lexical stress and sentence accents in phonetic input In phonetic input strings,lexical stress andsentence accents can be manually indicated by the user, by using a single quote ( ) or double quote ( ) respectively. Note that manually inserted lexical stress or sentence accents may have no effect in RealSpeak. The RealSpeak synthesis module may indeed have reasons to override the requested sentence accent, and thus not realize it. The Text-To-Speech system will automatically convert all lexical stress marks into sentence accents in case no manually added sentence accents are found in the phonetic input string. User's Guide for Castilian Spanish Chapter I/15

16 Chapter I Castilian Spanish Text-to-Speech system For Example: <ESC>/+mja. mi.go_se_ Ja.ma_ xwan*.<esc>/+. is the same as <ESC>/+mja. mi.go_se_ Ja.ma_ xwan*.<esc>/+ (Mi amigo se llama Juan.) If phonetic input contains at least one manually added sentence accent, no additional sentence accents are assigned by the text-to-speech system. Therefore, only those words marked with " will get a sentence accent. As a consequence, a message containing only one manual sentence accent will have an almost flat intonation on all the other words. For Example: <ESC>/+mja. mi.go_se_ Ja.ma_ xwan*.<esc>/+ (Only one sentence accent will be realized.) Phonetic input can also be combined with orthographic input. If no sentence accents are found in the input text (indicated by <ESC>" in orthographic input, or by " in phonetic input), the Text-To-Speech system will automatically assign sentence accents. In the orthographic part of the input, the Text-To-Speech system will realize these sentence accents on the basis of part-of-speech and syntactic information. In the phonetic part of the input, all lexical stress marks (if any) will be converted into sentence accents. If there are no lexical stress marks, no sentence accent will be realized for the phonetic part of the input (see point 1 above). If the user has manually specified one or more sentence accents, no additional sentence accents will be realized (see point 2 above). User's Guide for Castilian Spanish Chapter I/16

17 Castilian Spanish Text-to-Speech system Chapter I Si hace buen tiempo mañana, nos iremos a <ESC>/+man. xa.tan<esc>/+. (No sentence accents are found; The Text- To-Speech system will automatically assign sentence accents.) Si hace buen tiempo mañana, nos iremos a <ESC>/+man. xa.tan<esc>/+. (A sentence accent is specified in the phonetic part of the input text. No additional sentence accents will be realized.) Si hace buen tiempo <ESC> mañana, nos iremos a <ESC>/+man. xa.tan<esc>/+. (A sentence accent is specified in the orthographic part of the input text. No additional sentence accents will be realized. Hence, the lexical stress which is specified in the phonetic part will NOT be converted into a sentence accent.) Si hace buen tiempo <ESC> mañana, nos iremos a <ESC>/+man. xa.tan<esc>/+. (Two sentence accents are specified; no additional sentence accents will be realized.) The Castilian Spanish L&H+ and UNIPA Phonetic Alphabets L&H+ Symbol L&H+ Transcription Vowels UNIPA Symbol UNIPA Transcription As in: i 'bi.bo i 'bi.bo vivo e 'pe.so e 'pe.so peso a 'ka.sa a 'ka.sa casa o 'bo.bo o 'bo.bo bobo u ku.'n~a.do u ku.'n~a.do cuñado User's Guide for Castilian Spanish Chapter I/17

18 Castilian Spanish Text-to-Speech system Chapter I L&H+ Symbol L&H+ Transcription Consonants UNIPA Symbol UNIPA Transcription As in: p 'pe.pa p 'pe.pa pepa b 'kam.bjo b 'kam.bjo cambio t 'ti.a y 'ti.a tía d 'de.do d 'de.do dedo k 'ki.lo k 'ki.lo kilo g an.'gar6 g an.'gar6 hangar m 'a.mo m 'a.mo amo n 'ne.ne n 'ne.ne nene n~ 'ni.n o n~ 'ni.n o niño r 'ka.ro r 'ka.ro carro r6 'ka.r6o r6 'ka.r6o caro B 'bo.bo B 'bo.bo bobo f 'fin f 'fin fin T 'Te.r6o T 'Te.r6o cero D 'de.do D 'de.do dedo s 'se.so s 'se.so seso J 'ma.jo J 'ma.jo mayo x 'a.xo x 'a.xo ajo G 'a.gwa G 'a.gwa agua j 'Tje.Go j 'Tje.Go ciego w 'kwen.ta w 'kwen.ta cuenta l 'la.ta l 'la.ta lata t&s 't&si.ko t+s 't+si.ko chico d&z 'd&zo d+z 'd+zo Yo NOTE Note that the L&H+ alphabet is not SSML compliant. For SSML, use the UNIPA alphabet. User's Guide for Castilian Spanish Chapter I/18

19 Chapter I Castilian Spanish Text-to-Speech system Using a User Dictionary For information on how to create and use user dictionaries, please refer tothe User Configuration chapterin the RealSpeak Telecom Programmer s Guide. Using the Microsoft SAPI5 Lexicon User Lexicons Microsoft SAPI5 provides lexicons so that users and applications can specify pronunciation and part of speech information for particular words. As such, all SAPI compliant TTS engines should use these lexicons to guarantee uniformity of pronunciation and part of speech information. There are two types of lexicons in SAPI: user lexicons and application lexicons. Each user who logs in to a computer will have a User Lexicon. Initially, this lexicon is empty; words can be added either programmatically, or by using an engine's add/remove words UI component (for example, the sample application Dictation Pad provides an Add/Remove Words dialog). Application Lexicons Applications can create and ship their own lexicons of specialized words. These lexicons are fixed and cannot be edited. Detailed information on how to use the MS SAPI5 lexicons can be found in the manual Microsoft Speech SDK V5.1, chapter ISpLexicon Interface. User's Guide for Castilian Spanish Chapter I/19

20 Chapter I Castilian Spanish Text-to-Speech system The Castilian Spanish SAPI5 Phoneme List SAPI5 Symbol To add entries to the lexicon, the user should use a set of language specific phonemes. The language specific phoneme list for Castilian Spanish is given below. SAPI5 Symbols As in: SAPI5 Transcription SAPI Phone ID A valle V A JJ E 10 E pero P E R O 11 I pico P I K O 12 O toro T O R O 13 U duro D U R O 14 B barista, B A R I S T A, B I N O vino 18 B cabo K A B O 18 CH mucho M U CH O 21 D donde D O N D E 16 D nada N A D A 16 F fácil F A TH I L (F A S I L) 23 G gato G A T O 20 G luego L W E G O 20 JJ yerno, JJ E R N O 22 JJ caballo K A B A JJ O 22 J rey, pie RR E J, P J E 33 K casa K A S A 19 L lejos L E X O S 29 M mismo M I S M O 26 N nunca N U N K A 27 NJ año A NJ O 28 P padre P A D R E 17 R puro P U R O 31 RR torre, T O RR E, RR O X O rojo 32 S sala S A L A 24 T tomo T O M O 15 TH cinco TH I N K O 35 W deuda, D E W D A, M W I muy 34 X mujer M U X E R 25 User's Guide for Castilian Spanish Chapter I/20

21 Castilian Spanish Text-to-Speech system Chapter I SAPI5 Symbol SAPI5 Symbols Meaning or example - syllable boundary (hyphen) 1! Sentence terminator (exclamation mark) 2 & word boundary 3, Sentence terminator (comma) 4. Sentence terminator (period) 5? Sentence terminator (question mark) 6 _ Silence (underscore) 7 1 primary stress 8 2 secondary stress 9 SAPI Phone ID User's Guide for Castilian Spanish Chapter I/21

22 Chapter I Castilian Spanish Text-to-Speech system Notes on the Castilian Spanish Text-To-Speech System The Castilian Spanish Text-To-Speech system has been designed to allow a correct pronunciation of any input written according to the rules of Castilian Spanish orthography. The following cases, however, require special attention. Cardinal Numbers When no periods are used to separate groups of digits, cardinal numbers up to 8 digits are pronounced as full numbers. Digit strings of 9 or more digits are read one by one When periods are used to separate groups of digits, cardinal numbers up to 15 digits are pronounced correctly as full numbers Decimal Numbers When no periods are used to separate groups of digits, decimal numbers up to 8 digits before and 8 digits after the comma are pronounced correctly as full numbers , ,123 When periods are used to separate groups of digits, decimal numbers up to 15 digits before and 8 digits after the comma are pronounced as full numbers. User's Guide for Castilian Spanish Chapter I/22

23 Chapter I Castilian Spanish Text-to-Speech system , , Ordinal Numbers Cardinal numbers from 1 to 10 followed by the ordinal distinctives and ª will be pronounced as ordinal numbers. The numbers 1 and 3 may also be folowed by er, 1 by ra or ro. 1ro 2o 4ª 3er Roman Numbers Roman numbers in isolation are read as cardinal numbers. CXXII : ciento veintidós Roman numbers from 1 to 7 following or preceding a word beginning with an uppercase letter are read as ordinal numbers. Otherwise, they are read as cardinals. User's Guide for Castilian Spanish Chapter I/23

24 Chapter I Castilian Spanish Text-to-Speech system Carlos I Isabel II Enrique VIII Alfonso XII I Congreso IV Demostración Carlos primero Isabel segunda Enrique octavo Alfonso doce primer congreso cuarta demostración Bank Account Numbers To have bank account numbers correctly pronounced, use hyphens between groups of digits. Groups of digits up to 8 digits will be pronounced as full numbers. Larger groups will be read digit by digit Telephone Numbers In order to ensure a correct pronunciation of telephone numbers, it is recommended to use spaces or hyphens to separate groups of digits. (91) (93) User's Guide for Castilian Spanish Chapter I/24

25 Chapter I Castilian Spanish Text-to-Speech system Dates The following formats are supported: Day/Month/Year Day-Month-Year Day can have one or two digits between 1 and 31,Month can have one or two digits between 1 and 12, andyear can have two or four digits without periods. 12/12/ /12/ /01/2000 Time Indications The following formats are supported (12 hours and 24 hours clock): (1 or 2 digits):(2 digits) (1 or 2 digits).(2 digits) The digits may be followed by usual abbreviations ("h.", "H."). 14: h. 14:30 H. User's Guide for Castilian Spanish Chapter I/25

26 Chapter I Castilian Spanish Text-to-Speech system Currencies The Castilian Spanish Text-To-Speech system supports the currency symbols $ (dólar), (libra), ƒ (florín) and (euro). 1 $ un dólar 100 $ cien dólares 1 una libra 100 cien libras 1 ƒ un florín 100 ƒ cien florines 1 un euro 100 cien euros 1 un centavo 100 cien centavos Abbreviations The Castilian Spanish Text-To-Speech system contains a dictionary to handle the most frequent abbreviations, such as: etc. etcétera D. don dcha. derecha Dr. doctor ptas. pesetas ms minuto, minutos s. siglo User's Guide for Castilian Spanish Chapter I/26

27 Chapter I Castilian Spanish Text-to-Speech system Acronyms Forms such as EPTA, UNICEF, NIF, IPC (or N.I.F., I.P.C., etc.) are pronounced using a combination of rules and dictionaries. PRI IPC NIF is pronounced as a single word is spelled is expanded (as Número de Identificación Fiscal) User's Guide for Castilian Spanish Chapter I/27

28 Realspeak Telecom SDK Chapter II Preprocessor User s Guidefor Castilian Spanish V4.0

29 Chapter II Preprocessor Preprocessor Introduction The ScanSoft preprocessor (EMPP) has been developed to analyze a specific type of text: messages. messages differ from any average type of text in both structure and contents. An message consists of two clearly distinguished parts: the header and the body. A substantial part of the header contains routing and administrative information, which is irrelevant to the user. Both the header and the body contain all kinds of specific text features, e.g. addresses, emoticons such as smileys, etc. Furthermore, informal writing is often combined with a lack of grammatical conventions. Spelling rules are frequently violated, punctuation is often omitted, etc. Although the standard ScanSoft Text-To-Speech system can handle special text items (abbreviations, numbers, dates, etc.), it is not capable of correctly handling all specific text features. These text features are therefore dealt with by the preprocessor. The EMPP transforms specific information into a format that complies with the rules of the standard ScanSoft Text-To-Speech system. The EMPP is a plug-in preprocessing module of the ScanSoft Text-To-Speech system. It replaces the preprocessor of the standard Text-To-Speech system. In the following sections you will find a description of the functioning of the ScanSoft preprocessor as well as an overview of its features. The preprocessor has two main tasks: processing of the header and processing of the body of the message. The input to the EMPP consists of one or more messages. In order to process the header, the EMPP extracts relevant header fields and then provides an intelligent header field reading. User's Guide for Castilian Spanish Chapter II/29

30 Chapter II Preprocessor During the processing of the body, the text is divided into smaller text units, called text-to-speech messages, which are synthesized by the Text-To-Speech system. Text normalization is applied to specific text features such as addresses, proper names, emoticons, URLs (Universal Resource Locators), etc. For the text normalization of an message, the ScanSoft EMPP applies linguistic rules and performs dictionary look-up, in order to yield an adequate phonetic transcription. The EMPP also supports the ScanSoft user dictionary mechanism, which allows the user to customize the output of the processing. Header Processing Header Field Extraction An message consists of two clearly distinguished parts: the header and the body. The EMPP detects the header and extracts the relevant header fields. Information that is of no interest to the user (such as routing information) is not retained. The EMPP extracts the following header fields: From Field Date Field Subject Field Contains the sender s name and/or addres Contains the date and time of sending Optionally contains the subject of the The extraction of the header fields is based on the detection of specific keywords in the header. The supported keywords for the extraction of the header fields are listed below: From Field Date Field Subject Field: From: Author: Sender: De: Von: Date: Enviado: Gesendet: Subject: Subj: Asunto: Betreff: User's Guide for Castilian Spanish Chapter II/30

31 Chapter II Preprocessor The following is an example of header field extraction. The original header holds information that is irrelevant to the user. After extraction of date, sender and subject, the processed header merely mentions the Date field, the From field and the Subject field: Original header: From mundivia.es Mon Aug 04 11:37: Path: xenon.inbe.net!inbe.net!krypton.inbe.net! INbe.net! news.nl. innet.net! INnl.net!hunter.premie r.net From: es Newsgroups: soc.culture.spain Subject: Re: El voseo en Extremadura Date: Thu, 31 Jul :56: Organization: Netcom Lines: 30 Message-ID: ix.netcom.com> References: edu> netcom.com> netcom.com> Reply-To: ix.netcom.com NNTP- Posting- Host: mia- fl3-07.ix.netcom.com Mime- Version: 1.0 Content-Type: text/plain; charset=iso Content-Transfer- Encoding: 8bit X-NETCOM- Date: Thu Jul 31 6:54:48 AM CDT 1997 X-Mailer: Mozilla 3.0C- NC320 ( Win95; I) Xref: xenon.inbe. net soc.culture.spain: Status: N Extracted header fields: From: lmarina@mundivia.es Subject: Re: El voseo en Extremadura Date: Thu, 31 Jul :56: User's Guide for Castilian Spanish Chapter II/31

32 Chapter II Preprocessor Header Field Reading After the header fields have been extracted, they are processed by the EMPP. The header field keywords (see above) are replaced by an introductory message. The remainder of the header fields is processed by the EMPP in order to allow the Text-To-Speech system to intelligently read the fields. From Field The From field keyword is replaced by the introductory message Mensaje de:. From: Pablo Aguirre is pronounced: Mensaje de: Pablo Aguirre The remainder of the From field is further processed by the EMPP. The EMPP supports From fields that either consist of a) a proper name b) a proper name and an address c) an address a) - b) In case the From field contains a proper name, this name and only this name is sent to the Text-To-Speech system. This means that if both a name and an address are found in the From field, the address will not be read by the Text-To-Speech system. From: Mauricio de Pablo From: "Paco Manzano" <manzano@alpha.ee.upenn.edu> From: Marta Arribas at IepXchgPO are pronounced: Mensaje de: Mauricio de Pablo Mensaje de: Paco Manzano Mensaje de: Marta Arribas User's Guide for Castilian Spanish Chapter II/32

33 Chapter II Preprocessor c) In case the From field contains only an address, the EMPP extracts the name out of the address and reads the domain name literally. For Example: From: From: unizar.es!pablo From: From: rci.es!pozas at Internet are pronounced: Mensaje de: Alex arroba scansoft punto com Mensaje de: Pablo arroba unizar punto e s Mensaje de: t arroba b p e punto e s Mensaje de: pozas arroba r c i punto e s Date Field The Date field keyword is replaced by the introductory message Fecha: The Date field contains the date and time of sending. The EMPP supports multiple date and time formats, which are transformed to a uniform format that complies with the rules for date and time indications of the ScanSoft Text-To-Speech system. The EMPP only pronounces the date. The EMPP supports dates in the following formats: Date: 01/18/96 13:45:46 EDT Date: Thu, 7 Dec :45 AM Date: 13 Mar :55:27 GMT are pronounced: Fecha: 18 de Enero del 96 Fecha: Jueves, 7 de Diciembre de 1995 Fecha: 13 de Marzo de 1996 User's Guide for Castilian Spanish Chapter II/33

34 Chapter II Preprocessor Subject Field The Subject field keyword is replaced by the introductory message Asunto:. The Subject field can contain all kinds of data, but may also be empty. The EMPP searches for keywords that are typical of the subject field (e.g. RE, FYI, FW). Subject: RE: Regalo para Elena Subject: FYI: documento en Word 6 Subject: FWD: virus! are pronounced: Asunto: Respuesta a: Regalo para Elena Asunto: Para su información Asunto: (reenvío) virus! body processing Message Extraction The preprocessor splits the body of the message into text-to-speech messages. This is done on the basis of a number of criteria, such as punctuation, capitalization, layout, intelligent abbreviation handling, etc. User's Guide for Castilian Spanish Chapter II/34

35 Chapter II Preprocessor The following examples illustrate some criteria for splitting the text into text-to-speech messages: Using sentence final punctuation and capital letters Si tienes alguna duda sobre lo que hemos hablado, ponte en contacto conmigo. Mi dirección de es: Gracias. Using layout Estructura del tema: 1) Introducción 2) Desarrollo de los componentes 3) Sumario Using intelligent abbreviation handling Para dudas o preguntas, consúltese el manual del Dr. López Ibor User's Guide for Castilian Spanish Chapter II/35

36 Chapter II Preprocessor Text Normalization An message typically contains specific text features, such as addresses, URLs, file names, emoticons, etc. The EMPP transforms these specific features into a format that complies with the rules of the standard text normalization of the ScanSoft Text-To-Speech system. The following are examples of specific text normalization: Support for multiple address formats Gianluca Leoni/SSFT/IEP/BE plknapp@ucunix.san.uc.edu Sara Ghelli at IepXchgPO excalib!flagg!ucsd.edu!rambo Support for URLs (Universal Resource Locators) gopher://tecno.com/11 Support for file names ldb001.tse sysinfo.exe lipedu.xls Processing of emoticons :-* is pronounced huy :( is pronounced bubu Processing of overuse of punctuation Atención!!!!!!!: VIRUS DE INTERNET!!!!!!!! Tengo un problema! #%&#@$<! Quién me puede ayudar? becomes: ATENCIÓN!: VIRUS DE INTERNET! Tengo un problema! Quién me puede ayudar? User's Guide for Castilian Spanish Chapter II/36

37 Chapter II Preprocessor Normalization of lay-out lines (e.g. part of an signature). These sequences of identical characters are not pronounced: o 10 or more identical digits o a word consisting of 5 or more identical US-ASCII encoded letters of the modern Latin alphabet o a sequence of 3 or more identical characters that are no letters, no digits, no sentence-final punctuation marks ('. ', ' ', '? ', ' ', '! ') and no white spaces; e.g. '&', '#', '%', '*', '-' oooooooooooooooooooooooooooooo will be removed. Processing of inserted mail Marisa> Oye Emilio, cuánto hace que no llamas Marisa> a tu hermano para preguntarle como Marisa> está? Marisa> Si no te lo recuerdo yo, no lo haces. Emilio> Es verdad. Es que he tenido bastante Emilio> trabajo últimamente. Emilio> Ahora tiene una página en Internet. Emilio> Su dirección es becomes: Marisa: Oye Emilio, Cuánto hace que no llamas a tu hermano para preguntarle como está? Si no te lo recuerdo yo no lo haces.. Emilio: Es verdad. Es que he tenido bastante trabajo últimamente. Ahora tiene una página en Internet. Su dirección es Processing of Question/Answer (FAQ) P. Estimado Sr. González. Tengo en mi poder unos artículos sobre arquitectura de posguerra que quisiera que examinase. Podría enviarselos?. laut@unizar.tm.es R. Querido amigo. Como usted probablemente sabe, estoy preparando un coloquio sobre User's Guide for Castilian Spanish Chapter II/37

38 Chapter II Preprocessor este tema. No sólo me encantará leer sus artículos, sino que le invito muy cordialmente, si usted esta interesado, a tomar parte en este coloquio. becomes: Pregunta: Estimado Sr. González. Tengo en mi poder unos artículos sobre arquitectura de posguerra que quisiera que examinase. Podría enviarselos?. laut@unizar.tm.es Respuesta: Querido amigo. Como usted probablemente sabe, estoy preparando un coloquio sobre este tema. No sólo me encantará leer sus artículos, sino que le invito muy cordialmente, si usted esta interesado, a tomar parte en este coloquio. Language Specific Text Normalization The Preprocessor for Castilian Spanish is able to convert accented characters. In messages, accents and n-tilde can be transcribed in several ways: camio'n camion espanol espan~ol espanyol Accents or tildes may also be dropped: camion espanol The Preprocessor for Castilian Spanish correctly converts all possible ways of writing accented words. English words Since is an international medium, Castilian Spanish messages will inevitably contain many English words that might refer to Internet, electronic mail or soft- and hardware. The typical jargon is handled by the exceptions dictionary of the preprocessor. This dictionary is a lexicon for terminology and provides the TTS system with an adequate Castilian Spanish transcription for a number of English words. User's Guide for Castilian Spanish Chapter II/38

39 Chapter II Preprocessor Rlogin /+ '?er6.'lo.gin techweb /+ 'tex.'web uunet u.u. /+ 'net filemanager /+ 'fajl.'ma.na.jer6 projtech /+ 'pr6oj.'tek techtalk /+ 'tex.'tok abstract /+ '?abs.tr6akt access /+ '?ak.ses shareware /+ 'ser6.wer6 User's Guide for Castilian Spanish Chapter II/39

40 Chapter II Preprocessor Customizing the Preprocessor The preprocessor supports the standard ScanSoft Text- To-Speech SDK user dictionary mechanism, which allows the user to customize the output of the preprocessor. The user dictionary is consulted both during the header processing and the body processing. For more information on how to build and use user dictionaries, see the User Configuration chapter of theprogrammer s Guide. Customizing the Header The user dictionary is consulted during the header processing while reading the From field and the Subject field. From Field The From field either consists of a) a proper name b) a proper name and an address c) an address a) In case the From field contains a proper name only, the name is passed on to the user dictionary. If the lookup is successful, the proper name is substituted by the replacement string. If not, the name is further processed by the header reading module. User's Guide for Castilian Spanish Chapter II/40

41 Chapter II Preprocessor If the user dictionary contains the following line: John /+ 'd&zon the following From field: From: John Ferrari becomes: Mensaje de: <esc>/+ d&zon <esc>/+ Ferrari b) In case the From field contains a proper name and an address, the EMPP first passes the address to the user dictionary. If the lookup is successful, both the proper name and the address are substituted by the replacement string. If not, the EMPP passes the proper name to the user dictionary. If this lookup is successful, the name and the address are substituted by the replacement string. If not, the name is further processed by the header reading module. The address will not be read by the Text-To-Speech system. If the user dictionary contains the following lines: psmith@scansoft.com Pablo, mi tío de Indonesia Andrés un colega de mi hermana Schrijver /+?es.'xr6ej.ber6 the following From fields: From: psmith@scansoft.com (P. Smith) Author: andres@elis.rug.ac.be (andrés) From: "Alejandro De Schrijver" <DeSchrijver@ALPHA.ee.upenn.edu> User's Guide for Castilian Spanish Chapter II/41

42 Chapter II Preprocessor become: Mensaje de: Pablo, mi tío de Indonesia Mensaje de: un colega de mi hermana Mensaje de: Alejandro De <esc>/+?es.'xr6ej.ber6 <esc>/ c) In case the From field contains an address only, the complete address is looked up in the user dictionary. If the lookup is successful, a proper name is added to the From field. If not, only the domain part is sent to the user dictionary. The EMPP first calls the dictionary for the complete domain part. If the lookup is successful, the complete domain part is substituted by the replacement string. Otherwise, the EMPP cuts off the left most sublevel domain and repeats the lookup and matching procedure for the remainder of the domain part. If the lookup is successful, the remainder of the domain part is substituted by the replacement string. This procedure is called repeatedly until the top level domain is encountered. If none of the lookups is successful, the address is further processed by the header reading module. If the user dictionary contains the following lines: alberto.luna@arrakis.es mi mejor amigo telemir.es Veracruz (Figueras) the following From fields: Sender: alberto.luna@arrakis.es From: fernando.fuentes@telemir.es become: Mensaje de: mi mejor amigo Mensaje de: fernando heras en Veracruz (Figueras) User's Guide for Castilian Spanish Chapter II/42

43 Chapter II Preprocessor NOTE To allow a correct processing of the From field, the replacement string in the user dictionary should not contain an address. Subject Field Each word in the Subject field is sent to the user dictionary. If the lookup is successful, the replacement string is sent directly to the Text-To-Speech system. If not, the Subject field is further processed by the header reading module. If the user dictionary contains the following lines: EMI IDT /+ '?e.'e.me.'i I D T the following Subject fields: Subject: Jueves: Conferencia con el EMI Subject: sobre las prestaciones de IDT are pronounced: Asunto: Jueves: Conferencia con el <esc>/+'?e.'e.me.'i <esc>/+ Asunto: sobre las prestaciones de I D T User's Guide for Castilian Spanish Chapter II/43

44 Chapter II Preprocessor Customizing the Body When a user dictionary has been loaded, the EMPP will call the dictionary for every word of the body. If the word is found in the user dictionary, it is substituted by the replacement string. If not, the body is further processed by the body processing module. If the user dictionary contains the following line: msg mensaje the word "msg" in the following sentence: Tu msg me ha llegado hoy. Tengo que hacerte algunos comentarios. is replaced by the corresponding string found in the user dictionary: Tu mensaje me ha llegado hoy. Tengo que hacerte algunos comentarios User's Guide for Castilian Spanish Chapter II/44

45 Realspeak Telecom SDK Chapter III SSML Preprocessor User s Guide for Castilian Spanish V4.0

46 Chapter III SSML Preprocessor SSML Preprocessor Introduction SSML (Speech Synthesizer Markup Language) is part of a set of markup specifications by the W3C for voice browsers. General information regarding the RealSpeak SSML processor can be found in the SSML Support chapter of the Programmer s Guide. The RealSpeak Telecom SDK provides a built-in preprocessor that supports a large portion of the SSML 1.0 September 2004 Recommendation (REC). Moreover RealSpeak extends SSML with a number of Scansoft specific elements/attributes. The set supported by Scansoft is caled ScanSoft SSML (4SML). The section below describes language-specific SSML support included in the RealSpeak Telecom V4.0 Castilian Spanish language version. Castilian Spanish specific SSML markup XML encoding types for Castilian Spanish The encoding is specified in the XML text declaration ("<?xml?>") by the encoding declaration which is of the form encoding="<encodingname>". E.g. <?xml version="1.0" encoding="utf-8"?> RealSpeak Telecom V4.0 Castilian Spanish supports: Windows-1252 and ISO (ISO Latin1) The Unicode encoding UTF-8, UTF-16 and UCS-4 (Note that the alias "ISO UCS-4" is not supported) Any coding character set supported by the ICU component as long as the input text only contains characters that can be transcoded to the native coded character set, being Windows For more information about the character sets supported by ICU, take a look at the ICU website and User's Guide for Castilian Spanish Chapter III/46

47 Chapter III SSML Preprocessor NOTE 4SML Specifics for Castilian Spanish Encoding names are parsed case-insensitive; hyphens and underscores are ignored For reasons of compatibility with the standard Castilian Spanish system, the parallel text control sequence (<esc> sequence) is listed where applicable. As such, a similar TTS behavior can be created or combined with non-ssml driven text input. 4SML Tags Comment Corresponding control sequence High-level and document structure tags xml:lang Text normalization tags <say-as interpretas= xxx > <say-as interpretas= number format= cardinal > <say-as interpretas= number format= digits > <say-as interpretas= number format= decimal > <say-as interpretas= number > <say-as interpretas= number format= ordinal > <say-as interpretas= number format= telephone > es-es for Castilian Spanish. Attribute of speak, paragraph, sentence and voice. ; limited support in mode. In mode the only supported interpretas value is spel. <esc>\tn=number_cardinal\ <esc>\ tn=number_digits\ <esc>\ tn=number_decimal\ <esc>\ tn=number\ <esc>\ tn=number_ordinal\ <esc>\ tn=number_telephone\ User's Guide for Castilian Spanish Chapter III/47

48 Chapter III SSML Preprocessor 4SML Tags Comment Corresponding control sequence <say-as interpretas= number format= telephone detail= punctuation > <say-as interpretas= ordinal> <say-as interpretas= acronym > <say-as interpretas= acronym detail= strict > <say-as interpretas= measure > <say-as interpretas= letters > <say-as interpretas= letters detail= strict > <say-as interpretas= words > <say-as interpretas= date > <say-as interpretas= date format= mdy > <say-as interpretas= date format = dmy > <say-as interpretas= date format= ymd > <say-as interpretas= date format= ym > <say-as interpret-as date format= my > <say-as interpretas= date format= dm > <say-as interpretas= date format= md > <say-as interpretas= date format= y > <esc>\ tn=number_telephone_punctuation\ <esc>\ tn=ordinal\ <esc>\ tn=acronym\ <esc>\ tn=acronym_strict\ <esc>\ tn=measure\ <esc>\ tn=letters\ <esc>\ tn=letters_strict\ <esc>\ tn=words\ <esc>\ tn=date\ <esc>\ tn=date_mdy\ <esc>\ tn=date_dmy\ <esc>\ tn=date_ymd\ <esc>\ tn=date_ym\ <esc>\ tn=date_my\ <esc>\ tn=date_dm\ <esc>\ tn=date_md\ <esc>\ tn=date_y\ User's Guide for Castilian Spanish Chapter III/48

49 Chapter III SSML Preprocessor 4SML Tags Comment Corresponding control sequence <say-as interpretas= date format= m > <say-as interpretas= date format= d > <say-as interpretas= time > <say-as interpretas= time format= h > <say-as interpretas= time format= hm > <say-as interpret as= time format= hms > <say-as interpretas= duration format= hms > <say-as interpretas= duration format= hm > <say-as interpretas= duration format= ms > <say-as interpretas= duration format= h > <say-as interpretas= duration format= m > <say-as interpretas= duration format= s > <say-as interpretas= duration > <say-as interpretas= curency > <say-as interpretas= telephone > <say-as interpretas= telephone detail= punctuation > <say-as interpretas= address > <esc>\ tn=date_m\ <esc>\ tn=date_d\ <esc>\ tn=time\ <esc>\ tn=time_h\ <esc>\tn=time_hm\ <esc>\ tn=time_hms\ <esc>\ tn=duration_hms\ <esc>\ tn=duration_hm\ <esc>\ tn=duration_ms\ <esc>\ tn=duration_h\ <esc>\ tn=duration_m\ <esc>\ tn=duration_s\ <esc>\ tn=duration\ <esc>\ tn=currency\ <esc>\ tn=telephone\ <esc>\ tn=telephone_punctuation\ <esc>\ tn=address\ User's Guide for Castilian Spanish Chapter III/49

50 Chapter III SSML Preprocessor 4SML Tags Comment Corresponding control sequence <say-as interpretas= spell > <say-as interpretas= name > <say-as interpretas= net format= > <say-as interpretas= net format= uri > <say-as interpretas= net > Pronunciation tags <phoneme alphabet= unipa > See section the Castilian Spanish L&H+ and UNIPA phonetic alphabets for an overview of the alphabet. <esc>\ tn=spell\ <esc>\ tn=name\ <esc>\ tn=net_ \ <esc>\ tn=net_uri\ <esc>\ tn=net\ User's Guide for Castilian Spanish Chapter III/50

51 Realspeak Telecom SDK Chapter IV Custom G2P Dictionaries User s Guide for Castilian Spanish V4.0

52 Chapter IV Custom G2P Dictionaries Custom G2P Dictionaries Introduction ScanSoft's RealSpeak system now offers support for custom G2P dictionaries. A custom G2P dictionary module is an add-on module specifically designed to improve the quality of pronunciation for specific kinds of words. The Castilian Spanish system is currently not designed to support the use of a custom G2p dictionary module. User's Guide for Castilian Spanish Chapter IV/52

53 Realspeak Telecom SDK Appendices User s Guide for Castilian Spanish V4.0

54 Castilian Spanish voice names Appendix A Appendices Appendix A: Castilian Spanish voice names The Telecom RealSpeak /Host Text-To-Speech system now supports selecting the voice and language via a string as well as a define (please see the definition for the function TtsInitialize(Ex)() in the Programmers Guide and also thebackwards Compatibility Guide for details). The name strings for the currently supported Castilian Spanish voices are listed in the table below. Castilian Spanish Voice Name Strings Voice Isabel Name String Isabel The string to use to set the language to Castilian Spanish is Castilian Spanish User's Guide for Castilian Spanish Appendix A/54