JP2015121731A - Pronunciation dictionary conversion model creation device, pronunciation dictionary conversion device, method using the same, program, and storage medium for program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a pronunciation dictionary conversion model creation device for creating a pronunciation dictionary suitable for voice recognition task with a small amount of cost.SOLUTION: A pronunciation dictionary conversion label preparation part receives, as input, a set of original words constituting a word string and voice data of the original words, and pronunciation variation patterns, recognizes the voice data using a context-free grammar taking an acoustic model and the pronunciation variation patterns into account, and outputs as many pronunciation dictionary conversion labels as the number of the pronunciation variation patterns, the pronunciation dictionary conversion labels composed of a set of the original words and changed patterns corresponding to the pronunciation variation patterns. Next, a pronunciation dictionary conversion model learning part receives, as input, the pronunciation dictionary conversion labels and learns a pronunciation dictionary conversion model in which a conditional probability that converted words corresponding to changed patterns appear from the pronunciation dictionary conversion labels is modeled by machine learning.

Description

  The present invention relates to a pronunciation dictionary conversion model creation device, a pronunciation dictionary conversion device, a method thereof, a program, and a recording medium for creating a model for realizing pronunciation dictionary conversion corresponding to a target task.

  A general speech recognition system uses a language model for linguistic prediction and an acoustic model for acoustic prediction. Furthermore, a pronunciation dictionary representing “a relationship between words and their pronunciation” is used as a link between linguistic information and acoustic information. This pronunciation dictionary is often described as a recognition dictionary. The pronunciation dictionary is well known and is disclosed in Non-Patent Document 1, for example.

  The pronunciation dictionary is basically given the pronunciation of each word based on the regular reading kana by the morphological analyzer. However, in many cases, humans do not pronounce the words exactly as they are read. For example, the normal reading pseudonym of the word string “said” is “to” but may be pronounced “yutto”. Thus, the relationship between words and pronunciations is not a one-to-one relationship, and stochastic fluctuations can occur.

  If a pronunciation dictionary that captures such probabilistic variations in pronunciation can be constructed, an effect of improving speech recognition performance can be expected. However, it should be noted that the manner in which pronunciation changes occur varies greatly depending on the voice recognition task. For example, a teacher's utterance in an educational setting changes pronunciation in a polite direction. Specifically, a phenomenon in which the pronunciation becomes longer or a phenomenon in which an interval occurs during utterance occurs.

  On the other hand, in conversations with friends, pronunciation is not polite, and phenomena such as dropping of sounds during pronunciation are likely to occur. Therefore, it is important in speech recognition to construct an appropriate pronunciation dictionary for each assumed speech recognition task.

  In order to construct a pronunciation dictionary suitable for a speech recognition task, a method for predicting pronunciation variation from word notation and regular reading information for an arbitrary word has been proposed. The method is based on statistical learning. For statistical learning, learning data is required first. As learning data, a set of speech data and its word sequence is used.

  In the prior art, first, speech data is continuously recognized using only an acoustic model, and a phoneme sequence including pronunciation variation is obtained. At the same time, a regular phoneme sequence is obtained by performing morphological analysis on the word sequence. By this processing, data of “regular phoneme sequence-varied phoneme sequence” is constructed, and statistical modeling is performed from this data. A method using a decision tree is disclosed in Non-Patent Document 2, for example, and a method using a neural network or the like is disclosed in Non-Patent Document 3, for example.

Kiyohiro Shikano et al. “IT Text Speech Recognition System”, Ohm Publishing Co., pp.91-92. Riley, M., et al., “Stochastic pronunciation modeling from hand-labelled phonetic corpora”, Speech Communication, 29 (2-4): 209-224, November 1999. T.Fukada, T.Yoshimura, and Y.Sagisaka. Automatic generation of multiple pronunciations based on neural networks and language statistics.Speech Communication, 27: 63-73, 1999.

  Since the conventional pronunciation dictionary construction method uses the result of continuous speech recognition of speech data, there is a problem that a large amount of learning data is required to realize modeling of the phoneme level covering the range of pronunciation variation. . For every phoneme, change the range of these pronunciation variations, including “substitution variation (phoneme is replaced by another phoneme)”, “insertion variation (new phoneme is added)”, “dropping variation (phoneme disappears)” In order to realize the modeling of the covered phoneme level, a large amount of learning data must be prepared.

  The present invention has been made in view of this problem, a pronunciation dictionary conversion model creation device that creates a model for constructing a pronunciation dictionary with a small amount of learning data, a pronunciation dictionary conversion device that uses the model, and those It is an object of the present invention to provide a method, a program and a recording medium thereof.

  The pronunciation dictionary conversion model creation device of the present invention includes a pronunciation dictionary conversion label maintenance unit and a pronunciation dictionary conversion model learning unit. The pronunciation dictionary conversion label maintenance unit is a context-free grammar that takes the voice data from the acoustic model and the pronunciation variation pattern as an input, with the input word and voice data of the original word constituting the word series as input. Are used for voice recognition, and the pronunciation dictionary conversion label composed of a combination of the original word corresponding to the pronunciation variation pattern and the pattern after the variation is output for the number of the pronunciation variation patterns. The pronunciation dictionary conversion model learning unit receives the pronunciation dictionary conversion label as an input, and the pronunciation dictionary conversion model that models the conditional probability that the converted word corresponding to the changed pattern appears from the pronunciation dictionary conversion label by machine learning To learn.

  The pronunciation dictionary conversion device of the present invention includes a pronunciation dictionary conversion model, a pronunciation dictionary feature conversion unit, a pronunciation variation observation unit, and a pronunciation dictionary construction unit. The pronunciation dictionary conversion model is created by the above-described pronunciation dictionary conversion model creation device of the present invention. The pronunciation dictionary feature conversion unit receives a dictionary entry in the conversion source pronunciation dictionary to which only a regular reading pseudonym is assigned, and constructs a normal reading feature vector for the dictionary entry. The pronunciation variation observation unit obtains a probability value of each pronunciation variation pattern using the pronunciation dictionary conversion model with the normal reading feature vector as an input. The pronunciation dictionary construction unit arranges dictionary entries for each probability value of the pronunciation variation pattern and constructs a pronunciation dictionary in which pronunciation variation is considered

  According to the pronunciation dictionary conversion model creation device of the present invention, the speech data is speech-recognized using the acoustic model and the context-free grammar considering the pronunciation variation pattern in units of the speech data of the original words constituting the word series. A pronunciation dictionary conversion model is created from the result. Therefore, the pronunciation dictionary conversion model corresponding to the pronunciation variation pattern can be created with a small amount of data.

  Further, the pronunciation dictionary conversion device of the present invention uses a dictionary entry in the conversion source pronunciation dictionary to which only regular reading pseudonyms are given as a pronunciation dictionary in which pronunciation variation is considered using the above-described pronunciation dictionary conversion model of the present invention. Convert to Therefore, a pronunciation dictionary suitable for a speech recognition task can be constructed at a low cost.

The figure which shows the function structural example of the pronunciation dictionary conversion model creation apparatus 100 of this invention. The figure which shows the operation | movement flow of the pronunciation dictionary conversion model creation apparatus. The figure which shows an example of a context free grammar. The figure which shows the function structural example of the pronunciation dictionary conversion level maintenance part 110. FIG. The figure which shows the operation | movement flow of the pronunciation dictionary conversion level maintenance part 110. FIG. The figure which shows the function structural example of the pronunciation dictionary conversion model learning part 140. FIG. The figure which shows the operation | movement flow of the pronunciation dictionary conversion model learning part 140. FIG. The figure which shows the function structural example of the pronunciation dictionary converter 200 of this invention. The figure which shows the operation | movement flow of the pronunciation dictionary converter 200. The figure which shows the example of the dictionary entry of the conversion origin pronunciation dictionary. The figure which shows the example of the dictionary entry of the pronunciation dictionary in which the pronunciation variation converted by the pronunciation dictionary conversion apparatus 200 was considered.

  Embodiments of the present invention will be described below with reference to the drawings. The same reference numerals are given to the same components in a plurality of drawings, and the description will not be repeated.

  FIG. 1 shows a functional configuration example of the pronunciation dictionary conversion model creation device 100 of the present invention. The operation flow is shown in FIG. The pronunciation dictionary conversion model creation device 100 includes a pronunciation dictionary conversion label maintenance unit 110, an acoustic model 120, and a pronunciation dictionary conversion model learning unit 140. The pronunciation dictionary conversion model creation device 100 is realized by a predetermined program being read into a computer constituted by, for example, a ROM, a RAM, and a CPU, and the CPU executing the program. The same applies to other devices described below.

  The pronunciation dictionary conversion label maintenance unit 110 receives the original word constituting the word series and the voice data of the original word (the set 1, the set 2,..., The set M) and the pronunciation variation pattern as inputs, and the original word Voice data using the acoustic model 120 and a context-free grammar considering the pronunciation variation pattern, and the pronunciation dictionary conversion label consisting of a combination of the original word corresponding to the pronunciation variation pattern and the pattern after the variation is pronounced. The number of fluctuation patterns is output (step S110). The word series is, for example, a series of words “today”, “ha”, “sunny”, and “is”. For example, if a voice recognition task (teaching voice of a teacher's voice) is considered to be a careful pronunciation change, “put a pause for each mora”, “all mora sounds longer For example, there are three types such as “No change”.

  The context-free grammar is a grammar that allows only a pronunciation variation pattern, and is shown in FIG. 3, for example. For each of the original words that make up the word series “Today is Sunny”, words with a pronunciation variation pattern applied are arranged in series. Each arrow represents a selectable pronunciation path. Pronunciation dictionary conversion labels are output for the number of pairs of original words and audio data. In the example of the word series described above, four pronunciation dictionary conversion labels are output. A plurality of pronunciation dictionary conversion labels may be temporarily stored as the label group 130.

  The pronunciation dictionary conversion model learning unit 140 receives the pronunciation dictionary conversion label as an input, and the pronunciation dictionary conversion that models the conditional probability that the converted word corresponding to the changed pattern appears from the pronunciation dictionary conversion label by machine learning The model is learned (step S140). The pronunciation dictionary conversion labels are, for example, (today-insert a gap for each mora), (today-all mora sounds longer), (today-not changed). Three pronunciation dictionary conversion labels are input from the pronunciation dictionary conversion label maintenance unit 110 for each original word.

  The processing of the pronunciation dictionary conversion label maintenance unit 110 and the pronunciation dictionary conversion model learning unit 140 is repeated until all the combinations of the original word and the voice data of the original word constituting the word series are completed (No in step S150). . The control unit 150 controls the time series operation and the operation end control of the pronunciation dictionary conversion label maintenance unit 110 and the pronunciation dictionary conversion model learning unit 140. The function of the control unit 150 is not a special technical feature of this embodiment but a general one.

  Since the pronunciation dictionary conversion model learned with the above configuration is obtained by limiting the number of pronunciation variation patterns for each original word, it does not require a large amount of learning data for its creation. That is, the pronunciation dictionary conversion model can be learned with a small amount of data.

In the following, the operation of the pronunciation dictionary conversion model creation device 100 will be described in more detail by showing a more specific functional configuration example of each unit.
[Dictionary Dictionary Conversion Label Maintenance Department]
FIG. 4 shows a more specific functional configuration example of the pronunciation dictionary conversion label maintenance unit 110. The operation flow is shown in FIG. The pronunciation dictionary conversion label maintenance unit 110 includes a morphological analysis unit 111, a context free grammar construction unit 112, a context free grammar storage unit 113, a maximum likelihood sequence search unit 114, and a label generation unit 115.

  The morpheme analysis unit 111 obtains a morpheme analysis result with reading information from the word series (step S111). Any morphological analyzer can be used for the morphological analysis. For example, if the word sequence is "Today is sunny", the result of the morphological analysis is "Today; Kyo; Noun: Date: Continuum; is; Wa; Conjunctive particle; Sunny; Hare; Noun; Is; Death; : "End;" The morpheme analyzing means 111 may use any means as long as it satisfies the division of the word series into morphemes and the addition of regular readings.

  The context-free grammar construction unit 112 constructs a context-free grammar using the morpheme analysis result with reading information and the pronunciation variation pattern as input (step S112). The grammar allowed here is limited to a predetermined pronunciation variation pattern. This pronunciation variation pattern may be defined in various ways, but is limited to a pattern that can vary in units of words.

  Here, the variation pattern will be described as, for example, the above three types of “insert a pause for each mora”, “all mora sounds longer”, and “no change”. The morphological analysis result includes L words. In the above morphological analysis result, L = 4.

  In the case of the original word “Today”, “No change (as per the reading information)” means “Hare”, “Insert every mora” means “Ha, Les,” and “Mora all sounds longer” The pronunciation variation pattern is “Harley”.

  Considering these three types of pronunciation variation patterns for each word of the morphological analysis result, a context-free grammar that allows them is constructed (FIG. 3). Note that the pronunciation variation pattern is not limited to three types. For example, a pronunciation variation pattern such as “insert sound for each mora” may be added. In this case, there are four arrows indicating transitions between words shown in FIG. The constructed context-free grammar is stored in the context-free grammar storage unit 113.

ここでｔは入力音声の特徴ベクトル系列である。音声認識の場合、例えばメル周波数ケプストラム係数（ＭＦＣＣ）分析によって求めた特徴ベクトルを用いるのが一般的である。音響モデル１２０は、隠れマルコフモデルと混合ガウス分布とで規定される一般的なものである。単語系列と組の音声データを、例えば「キョウ ハー ハ、レ、デースー」とした場合、最尤系列は図３に太線で示すパスが選択される。 The maximum likelihood sequence search means 114 outputs a maximum likelihood sequence obtained by speech recognition of the original word speech data using the acoustic model 120 and the context free grammar stored in the context free grammar storage means 113 (step S114). The maximum likelihood sequence is to find a sequence s ^ having the maximum generation probability from the acoustic model in a path (route) permitted by the context free grammar. The sequence s ^ can be obtained by the following equation.

Here, t is a feature vector series of the input speech. In the case of speech recognition, it is common to use a feature vector obtained by, for example, mel frequency cepstrum coefficient (MFCC) analysis. The acoustic model 120 is a general one defined by a hidden Markov model and a mixed Gaussian distribution. When the speech data of the word series and the set is, for example, “Kyohaha, Les, Dasu”, the path indicated by the bold line in FIG. 3 is selected as the maximum likelihood series.

  The label generation means 115 receives the morpheme analysis result with reading information and the maximum likelihood sequence as input, and generates a pronunciation dictionary conversion label composed of the original word and the changed pattern (S115). When the maximum likelihood sequence is “Kyo ha ha, les, dessue”, the pronunciation dictionary conversion label is “Today; Kyo; ”All sounds longer”, “sunny; hare; nouns;-interleave every mora”, “is; death; judgment: end;-all mora sounds longer” as output Can be obtained.

[Dictionary Dictionary Conversion Model Learning Department]
FIG. 6 shows a functional configuration example of the pronunciation dictionary conversion model learning unit 140. The operation flow is shown in FIG. The pronunciation dictionary conversion model learning unit 140 includes a feature vector extraction unit 141 and a pronunciation dictionary conversion device model parameter learning unit 142.

  The feature vector extraction means 141 receives the pronunciation dictionary conversion label as an input, and compares the feature vector extracted from the word information of the original word constituting the pronunciation dictionary conversion label with the changed pattern constituting the pronunciation dictionary conversion label. The learned learning label is output (step S141). The learning label is expressed in the form of a feature vector x and an output label y.

  The feature vector extracting means 141 first extracts a feature vector from the word information of the original word. Various feature vectors can be used. For example, it is assumed that part-of-speech information and word mora length are used. If there are five types of parts of speech, such as “verb, noun, adjective, adverb, and others”, five dimensions are secured only with the part of speech information.

  If the original word is a noun, only the “noun” portion is set to 1 and the other portions are reset to 0. For example, the feature vector x = [0, 1, 0, 0, 0].

  Similarly, when considering the three types of “1 mora, 2 mora to 4 mora, 5 mora or more” for the mora length, if the original word is 1 mora, only that portion is set to 1 and the other portions 0 will be inserted in For example, the feature vector x = [0, 1, 0, 0, 0, 1, 0, 0]. The first five parts of speech information and the subsequent three parts represent mora length information.

  The output label y is, for example, 1 for “according to reading information”, 2 for “insert a pause (pause) for each mora”, and 3 for “all mora sounds longer”.

  If the pronunciation dictionary conversion label is “today; Kyo; noun: date and time-reading information as it is”, the learning label is “x = [0,1,0,0,0,0,1,0], y = 1. "

  The pronunciation dictionary conversion device model parameter learning unit 142 learns a pronunciation dictionary conversion model which is a model parameter for obtaining a conditional probability that the learning label is input, the feature vector is an input feature vector, and the output label y is output (step) S142). The model to be learned can model the conditional probability P (y | x) from the input feature vector x and the output label y. There are various types of modeling. For example, modeling is performed using a maximum entropy model.

ここでｗはモデルパラメータである。具体的な学習方法は、例えば参考文献１（北健二「言語と計算-４確率的言語モデル」東京大学出版会,pp.162-165）に記載された周知の方法を用いる。 The maximum entropy model is a model equivalent to a log-linear model and is a well-known model. The maximum entropy model can be expressed as

Here, w is a model parameter. As a specific learning method, for example, a known method described in Reference Document 1 (Kenji Kita “Language and Computation—4 Stochastic Language Model”, University of Tokyo Press, pp. 162-165) is used.

  When learning of the pronunciation dictionary conversion model is completed, a conditional probability P (1 | x that is according to the reading information is input by inputting a feature vector x configured by a pronunciation dictionary feature unit 220 of the pronunciation dictionary conversion device 200 described later. ), A conditional probability P (2 | x) that puts a gap for each mora, and a conditional probability P (3 | x) that all mora lengthens.

[Pronunciation dictionary converter]
FIG. 8 shows a functional configuration example of the pronunciation dictionary conversion apparatus 200 of the present invention. The operation flow is shown in FIG. The pronunciation dictionary conversion apparatus 200 includes a pronunciation dictionary conversion model 210, a pronunciation dictionary feature conversion unit 220, a pronunciation variation observation unit 230, and a pronunciation dictionary construction unit 240.

  The pronunciation dictionary conversion model 210 is a conversion model created by the pronunciation dictionary conversion model creation apparatus 100 described above. The pronunciation dictionary activation unit 220 receives the dictionary entry in the conversion source pronunciation dictionary to which only the regular reading kana is given as input, and constructs a normal reading feature vector for the dictionary entry (step S220).

  FIG. 10 shows an example of a dictionary entry of an existing pronunciation dictionary. The dictionary entry is one line in FIG. 10 and is composed of a normal reading of each word “cloudy”, part-of-speech information, a reading pseudonym, and a probability value thereof. The normal reading feature vector for the dictionary entry is generated according to the same rule as the feature vector extracting unit 141 described above. For example, the normal reading feature vector for the dictionary entry “cloudy; spider; noun; ⇒ spider = 1.0” is expressed in the form x = [0, 1, 0, 0, 0, 0, 1, 0]. . The five parts of speech information from the beginning indicates a noun, and the subsequent three mora length information indicates 2 mora.

  The pronunciation variation observation unit 230 receives the normal reading feature vector as an input and obtains a conditional probability P (y | x) of each pronunciation variation pattern using the pronunciation dictionary conversion model 210 (step S230). For example, if the normal reading feature vector x = [0, 1, 0, 0, 0, 0, 1, 0] of “cloudy; spider; noun; ⇒ spider = 1.0” is input, P (1 | x ) = 0.65, P (2 | x) = 0.23, and P (3 | x) = 0.12. The pronunciation variation observing unit 230 calculates the conditional probability according to the equation (2).

  The pronunciation dictionary construction unit 240 constructs a pronunciation dictionary in which pronunciation variation is considered by arranging dictionary entries for each conditional probability P (y | x) of the pronunciation variation pattern (step S240). The processing of the pronunciation dictionary activation unit 220, the pronunciation variation observation unit 230, and the pronunciation dictionary construction unit 240 is repeated until all dictionary entries are completed (No in step S250).

  FIG. 11 shows an example of a dictionary entry of the pronunciation dictionary that takes into account the pronunciation variation converted by the pronunciation dictionary conversion apparatus 200. In this example, three variation patterns are arranged for each word.

  According to the pronunciation dictionary conversion apparatus 200, the dictionary entry in the conversion source pronunciation dictionary to which only the regular reading pseudonym is given is considered for the pronunciation variation by using the pronunciation dictionary conversion model of the present invention in which the pronunciation variation is limited. To the pronunciation dictionary. Therefore, a pronunciation dictionary suitable for a speech recognition task can be constructed at a low cost.

  In the above-described embodiment, the feature vector x has been described using an example in which the feature vector x is represented by five parts of speech information and three mora length information. “Presence / absence of kanji” or “difference between notation and standard reading” may be added to the vector element. Further, an example using the maximum entropy model has been described as an example of machine learning, but any model can be used for modeling the conditional probability P (y | x). For example, a neural network may be used. In this case, the intermediate layer of the neural network is a sigmoid function and the output layer is a soft matrix function.

  As described above, according to the pronunciation dictionary conversion model creation device 100 of the present invention, continuous speech recognition as in the prior art is not performed, and a wide range of context free grammars assuming limited pronunciation fluctuations are used. There is no need to deal with fluctuations in pronunciation. As a result, a robust pronunciation dictionary conversion model can be created with a small amount of data.

  Further, the pronunciation dictionary conversion apparatus 200 of the present invention uses the pronunciation dictionary conversion model to construct a pronunciation dictionary suitable for the speech recognition task, so that it is possible to create a pronunciation dictionary adapted to the speech recognition task at low cost. . This pronunciation dictionary is also suitable for use in speech synthesis that reproduces the habits of individual speaking.

  When the processing means in the above apparatus is realized by a computer, the processing contents of the functions that each apparatus should have are described by a program. Then, by executing this program on the computer, the processing means in each apparatus is realized on the computer.

  The program is distributed by selling, transferring, or lending a portable recording medium such as a DVD or CD-ROM in which the program is recorded. Further, the program may be distributed by storing the program in a recording device of a server computer and transferring the program from the server computer to another computer via a network.

  Each means may be configured by executing a predetermined program on a computer, or at least a part of these processing contents may be realized by hardware.

  The present invention can be used in both the fields of speech recognition and speech synthesis.

Claims (8)

Speech recognition using the speech model and the context-free grammar considering the phonetic variation pattern, using the original word constituting the word sequence and the voice data of the original word and the phonetic variation pattern as input A pronunciation dictionary conversion label maintenance unit that outputs a pronunciation dictionary conversion label consisting of a combination of the original word corresponding to the pronunciation variation pattern and a pattern after variation, by the number of the pronunciation variation patterns;
Pronunciation dictionary conversion that learns a pronunciation dictionary conversion model in which the conditional probability that a converted word corresponding to the changed pattern appears from the pronunciation dictionary conversion label is modeled by machine learning from the pronunciation dictionary conversion label A model learning unit;
A pronunciation dictionary conversion model creation device comprising: In the pronunciation dictionary conversion model creation device according to claim 1,
The pronunciation dictionary conversion label maintenance department
From the above word series, morpheme analysis means for obtaining a morpheme analysis result with reading information;
Context-free grammar construction means for constructing a context-free grammar using the morpheme analysis result with reading information and the pronunciation variation pattern as input,
A context-free grammar storage means for storing the context-free grammar;
Maximum likelihood sequence search means for outputting a maximum likelihood sequence obtained by speech recognition of the original word speech data using an acoustic model and the context-free grammar;
Label generation means for generating a pronunciation dictionary conversion label composed of the original word and the pattern after variation, using the morphological analysis result with reading information and the maximum likelihood sequence as inputs,
A pronunciation dictionary conversion model creation device characterized by comprising: In the pronunciation dictionary conversion model creation device according to claim 1 or 2,
The pronunciation dictionary conversion model learning unit is
Using the pronunciation dictionary conversion label as an input, a feature vector is extracted from the word information of the original word constituting the pronunciation dictionary conversion label, the feature vector is used as an input feature vector, and the pattern after variation is used as an output label. Feature vector extraction means for outputting as a learning label,
A pronunciation dictionary conversion device that models a pronunciation dictionary conversion model that is a model parameter for obtaining a conditional probability that the output label is output using the feature vector as an input feature vector with the learning label as an input, using a maximum entropy model Model parameter learning means;
A pronunciation dictionary conversion model creation device characterized by comprising: A pronunciation dictionary conversion model created by the pronunciation dictionary conversion model creation device according to claim 1;
A dictionary entry feature generating unit that constructs a normal reading feature vector for the dictionary entry, using a dictionary entry in the conversion source pronunciation dictionary to which only a regular reading pseudonym is assigned,
Using the normal reading feature vector as an input, the pronunciation variation observation unit for obtaining the probability value of each pronunciation variation pattern using the pronunciation dictionary conversion model,
A pronunciation dictionary construction unit that constructs a pronunciation dictionary in which pronunciation variation is considered by arranging dictionary entries for each probability value of the pronunciation variation pattern;
A pronunciation dictionary conversion device comprising: The pronunciation dictionary conversion label maintenance unit receives a combination of the original word constituting the word series and the voice data of the original word and the pronunciation variation pattern, and the speech data is considered in the context of the acoustic model and the pronunciation variation pattern. Speech dictionary conversion using free grammar, and pronunciation dictionary conversion label that outputs pronunciation dictionary conversion labels consisting of pairs of the original word and the changed pattern corresponding to the pronunciation variation pattern for the number of the pronunciation variation patterns Process,
A pronunciation dictionary in which the pronunciation dictionary conversion model learning unit models the conditional probability that a converted word corresponding to the changed pattern appears from the pronunciation dictionary conversion label using the pronunciation dictionary conversion label as an input by machine learning Pronunciation dictionary conversion model learning process for learning conversion model,
A pronunciation dictionary conversion model creation method comprising: A pronunciation dictionary feature generation process, wherein a dictionary entry in the conversion source pronunciation dictionary to which only a regular reading pseudonym is given is input, and a pronunciation dictionary feature conversion process that constructs a normal reading feature vector for the dictionary entry;
A pronunciation variation observing process in which a pronunciation variation observation unit obtains a probability value of each pronunciation variation pattern using the pronunciation dictionary conversion model created by the pronunciation dictionary conversion model creation method according to claim 5 by using the normal reading feature vector as an input; ,
A pronunciation dictionary construction process in which a pronunciation dictionary construction unit constructs a pronunciation dictionary in which pronunciation variation is considered by arranging dictionary entries for each probability value of the pronunciation variation pattern.
A pronunciation dictionary conversion method comprising:   The program for making a computer perform the function of each part of the pronunciation dictionary conversion model creation apparatus in any one of Claims 1 thru | or 3, and the pronunciation dictionary conversion apparatus in Claim 4.   A computer-readable recording medium on which any one of the programs according to claim 7 is recorded.

Images