JPH07199992A - Information offering device - Google Patents

Abstract

PURPOSE:To provide the information offering device which is superior in quick reporting performance, enables a user to select a sentence to be read aloud next and offers the selected sentence with a speech that is easy for the user to hear. CONSTITUTION:This information offering device has a language processing part 1 which converts a text message wherein a message to be offered is in character notation into a text message to be read aloud through language processing and stores the message in a message storage part 2 while dividing it into sentences, and also stores the storage addresses of the respective sentences in a position information storage part 3, and a control part 4 which stores the message number and sentence number of the text message and sentence being read aloud in a work area of an internal RAM, determines the message number and sentence number of a message and a sentence to be read aloud next according to a user's indication supplied from a signal reception part 5 by referring to the position information storage part 3, and supplies them from the message storage part 2 to the speech synthesis part 6.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention accumulates a text message in which a message to be provided is written, converts it into voice in accordance with a user's instruction, and converts the voice into
For example, the present invention relates to an information providing device that provides a user with a telephone.

[0002]

2. Description of the Related Art Conventionally, a recording type information providing service has been provided as a form of a voice information providing service using a telephone or the like. What is this recording type information service?
The announcer (reader) records a voice message obtained by reading the manuscript in which the message is read on the information providing device, and when the user selects a desired service according to the prompt message reproduced by the information providing device, the selected content is selected. The corresponding voice message is provided to the user via the telephone.

[0003]

As described above, in the recording type information providing service, it is necessary to record the message to be provided. Since this work takes time depending on the message to be provided, the recording type information providing service provides highly urgent information such as breaking news of serious events, and the ever-changing sports match progress (results). It had the drawback that it was not suitable for providing information.

A message to be provided is usually composed of a plurality of sentences, but in a recording type information providing service, it is reproduced in message units. Therefore, there is also a problem of lack of interactivity that the user cannot move to the step of selecting the next message unless he / she listens to the selected voice message to the end. Furthermore, there is a drawback that an announcer who reads out the manuscript is required, and labor costs are high in providing the service.

On the other hand, in the information providing apparatus for regularly synthesizing voice from a conventional text message and outputting it,
Since it is extremely difficult to prepare a dictionary with sufficient ability, “reading mistakes” occur and the intonation and accent are unnatural, so they have not reached the level for practical use. However, in services such as inquiries about bank deposit balances, when the user repeats the account number represented by the PB signal (signal corresponding to the push button on the telephone) input by the user using the telephone, the above-mentioned rule composition is performed. An information providing device using is used. This is because the entered account number is information that the information providing device cannot prepare in advance, and the use of the information providing device using rule composition is indispensable for providing the service. Limited only if.

As described above, each of the conventional recording type information providing service and the information providing apparatus using rule composition has disadvantages. In addition, information such as the above breaking news and sports match progress (results) information, etc. can be obtained from newspaper reporters using the easily available telephones such as home and office telephones and public telephones on the street. There is a long-awaited information providing service that can provide a user with a message immediately.

The present invention has been made in view of the above-described circumstances, has an excellent promptness, allows the user to select a sentence to be read next, and makes it easy for the user to hear the selected sentence. It is an object to provide an information providing device that can be provided by voice.

[0008]

An information providing apparatus according to the present invention includes a language processing means for performing language processing on a text message in which a message to be provided is written and converting the text message into a text message for reading, and the text message for reading. Message storage means for dividing a plurality of sentences into sentence units and storing the same, position information storage means for storing the storage position of the sentence stored in the message storage means, and signal receiving means for receiving an instruction signal indicating the user's instruction. And holding information about the sentence being read, referring to the position information storage means, and determining a sentence to be read next according to the instruction of the user represented by the instruction signal received by the signal receiving means. Control means for controlling, and controlled by the control means,
And a voice synthesizing unit for converting the sentence to be read next stored in the message storing unit into a voice.

[0009]

According to the above structure, the language processing means stores the reading text message obtained by performing the language processing on the message to be provided in the message storage means in sentence units, and stores the storage position of each stored sentence. It is stored in the position information storage means. Further, the control means refers to the position information storage means, and causes the message storage means to supply the next speech to be read to the voice synthesis means in response to the instruction signal received by the signal receiving device, and outputs the voice. Therefore,
In addition to having excellent promptness, the user transmits a desired instruction signal to select a sentence to be read next, and the selected sentence is provided as a language-processed and easy-to-listen voice.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of an information providing apparatus according to an embodiment of the present invention. A language processing unit 1 receives a text message issued by a newspaper reporter or the like from the outside (flash news article system) and performs language processing on the text message to convert it into a text message for reading. Then, the text message for reading is stored, for example, in the message storage unit 2 configured by a RAM having a sufficient storage capacity, and its storage address (storage position) is set to, for example, R of a sufficient storage capacity.
It is stored in the position information storage unit 3 composed of AM.

Here, the contents stored in the position information storage unit 3 will be described with reference to FIG. FIG. 2 is a conceptual diagram for explaining a configuration of various information stored in the position information storage unit 3. As shown in this figure, one reading text message is composed of a plurality of sentences. For example, a reading text message having a message number of i (i is a natural number) is composed of j (j is a natural number) sentences.

In the position information storage unit 3, in addition to the message number and the number of sentences in the message, sentence numbers unique to each sentence (s11, s12, s13, ..., Si1, si2, ...
, Sij) and the storage address (add11, storage address) in the message storage unit 2 in which the sentence represented by each sentence number is stored.
add12, add13, ..., addi1, addi
2, ..., Addij) are stored in association with each other.

As a natural language processing technique for realizing the language processing unit 1 described above, here, "Japanese Correction Support System (REVISE)" (Research Practical Report, Vol. 36, No. 9)
No., PP.1159 to PP.1167). By using this "Japanese correction support system (REVISE)" in the first stage of rule composition, it prevents "misreading".
It is possible to achieve vocalization with correct intonation.

In FIG. 1, 4 is a control unit, 5 is a signal receiving unit, and 6 is a voice synthesizing unit. Each of the units 4 to 6 will be described below. The signal receiving unit 5 receives an instruction signal (P signal transmitted from a PB signal voice generator in the telephone operated by the user).
B signal) and supplies the corresponding signal to the control unit 4.

Here, the instruction signal is a signal representing a process requested by the user. For example, the instruction signal transmitted by pressing the button "7" of the telephone by the user is an unheard portion of the message being read. "Skip", the instruction signal sent by pressing the button "8" is the instruction signal sent by pressing the button "9" and "Back" to read again from the beginning of the sentence being read or the beginning of the previous sentence. Means "repeat" of the message being read.

The control unit 4 includes, for example, a CPU and an R (not shown).
OM, RAM, various I / O interfaces, etc., and manages message numbers and sentence numbers given to reading text messages and sentences corresponding to the message being read, and the sentence number of the sentence currently being read,
The sentence number of the sentence to be read next is determined according to the signal (indicating the user's instruction) supplied from the signal receiving unit 5. Further, when the sentence currently being read is finished or when a signal is supplied from the signal receiving unit 5, the control unit 4 refers to the position information storage unit 3 to store the sentence to be read next. After obtaining the position information, the reading text message stored in the storage position represented by the storage position information is supplied from the message storage unit 2 to the voice synthesis unit 6.

The voice synthesis unit 6 performs rule synthesis on the text message for reading supplied from the message storage unit 2 and outputs the voice. The voice signal output by the voice synthesizer 6 is output from the handset of the user's telephone. In addition, the user transmits an instruction signal from the telephone while a certain text message for reading is being read,
An instruction to change the reading order is given for each reading text message or for each sentence that composes the reading text message.

The operation of the information providing apparatus having such a configuration will be described with reference to FIG. FIG. 3 is a process diagram showing the operation of the reading process performed by the information providing apparatus according to the embodiment of the present invention. The operation shown in this figure starts from step S1 when the user accesses the information providing apparatus via the telephone. It is executed in order.

First, step S1 is a process performed immediately after the user accesses the information providing apparatus via the telephone, and the information providing apparatus provides a message that can be provided and an operation necessary for reading the message. Guide the contents. Here, when the user selects a desired message and performs an operation corresponding to the message on the telephone,
An instruction signal (PB signal) corresponding to the operation content of the user is transmitted from the telephone. Then, the process proceeds to step S2.

In step S2, the signal receiving section 5 receives the instruction signal transmitted in step S1 and supplies the corresponding signal to the control section 4. Then, the control unit 4 changes the message number and the sentence number of the message being read to the internal R
Store in the work area of AM. Next, in step S3,
Based on the signal supplied from the signal receiving unit 5 and the message number and the sentence number stored in the work area, the control unit 4 determines the message number of the reading text message to be read next and the sentence number of the sentence. decide. Then, the storage address stored in association with the message number and the sentence number is read from the position information storage unit 3,
The sentences composing the text message for reading stored in the storage address of the message storage unit 2 are sequentially supplied to the voice synthesis unit 6 to output the voice.

Step S4 is a process performed in parallel with the voice output in step S3, and while the user listens to the voice synthesis output of the desired message, he / she will be willing to listen to the message continuously. If you want to hear the text message again from the beginning, or if you want to hear the part of the text you are reading aloud, you can
Send instructions via phone to change the message or sentence to be read.

Step S5 is a process performed after step S4, and whether or not the instruction supplied to the control unit 4 via the telephone and the signal receiving unit 5 means stop reading aloud. To judge. The result of this judgment is "Yes".
If so, the reading process ends. That is, the information providing service ends. On the contrary, if the determination result is “No”, the process proceeds to step S6.

In step S6, the control unit 4 receives the instruction supplied in step S4 and the message number and sentence number information of the reading text message and sentence currently read in the work area of the internal RAM. The message number and sentence number of the next text message and sentence to be read aloud are determined based on the. Then, the storage address stored in the position information storage unit 3 is acquired in association with the determined message number and sentence number. Then, the process returns to step S3. In this way, the reading process is performed.

As described above, since the information providing apparatus according to the embodiment of the present invention performs the voice synthesis processing using the natural language processing and the rule synthesis, it provides excellent promptness and easy-to-listen voice output. Can be compatible. Also,
Since the sentence number and the storage address of the sentence forming the text message for reading are stored in the position information storage unit 3, the control unit 4 next responds to each instruction of “skip”, “back”, and “repeat”. You can change the sentence read out. Therefore, it is possible to provide an information providing service with excellent interactiveness.

[0025]

As described above, according to the present invention,
The word processing means stores a reading text message obtained by subjecting the message to be provided to language processing in the message storage means on a sentence-by-sentence basis, and stores the storage position of the stored sentence in the position information storage means. Further, the control means refers to the position information storage means, and causes the message storage means to supply the next speech to be read to the voice synthesis means in response to the instruction signal received by the signal receiving device, and outputs the voice. Therefore, in addition to having excellent promptness,
The user can select a sentence to be read next by transmitting a desired instruction signal, and the selected sentence is provided as a language-processed and easy-to-hear voice.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an information providing device according to an embodiment of the present invention.

FIG. 2 is a conceptual diagram for explaining a configuration of various information stored in a position information storage unit 3 of the present embodiment.

FIG. 3 is a process diagram showing an operation of a reading process performed by the information providing apparatus according to the present embodiment.

[Explanation of symbols]

 1 language processing unit 2 message storage unit 3 position information storage unit 4 control unit 5 signal receiving unit 6 speech synthesis unit

Claims (1)

[Claims] 1. A linguistic processing means for performing linguistic processing on a text message in which a message to be provided is written and converting it into a text message for reading, and a message storage for storing the text message for reading divided into sentence units. Means, position information storage means for storing the storage location of the sentence stored in the message storage means, signal receiving means for receiving an instruction signal indicating a user's instruction, and information for the sentence being read aloud. A control unit that refers to the position information storage unit and determines a sentence to be read next in accordance with the user's instruction represented by the instruction signal received by the signal receiving unit; A voice synthesizing unit for converting the next sentence read aloud stored in the message storing unit into a voice. That the information providing apparatus.