JP3787623B2 - Conversation expression generation device and conversation expression generation program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a conversation expression generating apparatus and a program therefor that can suitably generate a conversational expression based on a monologous expression.
[0002]
[Prior art]
At present, some of the means for providing information via communication means such as the Internet are accompanied by images, but text communication such as texts in web pages, e-mails, chats, and electronic bulletin boards is mainly used. It is. In particular, it can be said that many of the sentences in the Web page are monologous sentences unilaterally described by a single writer. In addition, in text communication such as e-mail, the dialogue (conversation) adjustment in which the other person interposes a question during the description (speaking) of one writer (speaker) is a face-to-face conversation in which they face each other in reality. Therefore, it is considered that each utterance is generally long and has a monologue property that is fragmented for each utterance. As described above, there is a large amount of information stored as monologous sentences on the Internet. However, monologous sentences are not a means for presenting information to everyone, and conversational sentences that express important parts in question-answer format are generally more familiar than specialized monologue sentences. The degree of understanding also tends to be high. Here, conversational sentences, such as interview articles and TV program scripts, are sentences in which conversations are stacked. Conversational sentences are compared to monologue sentences. Although it is difficult to make an objective and strictly structured description, it has the advantage of being able to present concise information focusing on important parts, Since it is the most common information exchange means used for secure communication, it is characterized by being more familiar than monologous sentences and being easy to understand the conversation partner and listener. Explaining this by taking a TV broadcast news program as an example, it is more familiar to introduce two or more personalities in a conversational format than an announcer reads a news manuscript alone on a daily topic. .
[0003]
As a method of using a conversational expression, for example, the same character agents corresponding to the user's Web browsing are allowed to perform a conversation or a skit related to the page, so that the user can consistently understand the Web page. Attempts have been made to obtain properties such as Agneta & Frida (Non-patent Document 1). In addition, based on soccer game information, Personality An agent (Gerd & Matze) that generates a conversation according to the above is also considered (Non-patent Document 2). Furthermore, an agent salon that allows conversation between agents generated by applying simple rules to the personal information of exhibition participants is also considered (Non-patent Document 3).
[0004]
[Non-Patent Document 1]
Kristina Hook (H "o" ok, K.) and three other authors, "Handling of a hidden Lutheran view in the interface: Dealing with the Lurking Lutheran view on Interfaces: Evaluation of the Agneta and Frida system ”, (Sitges, Spain), the workshop Behavior Planning for Life-Like Synthetic Characters”, 1999, p125-136.
[Non-Patent Document 2]
Elizabeth Andre (Er.), 1 other author, “Performance Through Performance: On the Use of Multiple Lifelike Characters in Knowledge-Based Presentation Systems ”, (USA), The Second International Conference on Intelligent User Interfaces (IUI2000), 2000, p. 1-8
[Non-Patent Document 3]
Kaku, Mase, “Agent Salon: Encourage Encounters and Dialogues Using Chat between Personal Agents” IEICE Transactions, Vol. 84, DI, No. 8, 2001, p1231-1243
[0005]
[Problems to be solved by the invention]
However, the content of utterances used in Agneta & Frida is determined in advance, and does not newly generate a conversational sentence. In Gerd & Matze, utterances are in response to discrete events related to ball movement in soccer such as passes and intercepts, and conversational expressions are expressed from monologous sentences that describe unrestricted events in the field. It cannot be generated. Furthermore, since items used for conversation in the agent salon are discrete items such as individual visit histories and evaluation data for exhibitions, they correspond only to limited events as in the case of Gerd & Matze.
[0006]
Therefore, in any conventional mode, it is impossible to generate a conversation expression that is easy to understand and friendly based on an arbitrary monologous sentence.
[0007]
In view of the above problems, the main object of the present invention is to be able to convert an arbitrary monologous sentence that is not limited in the field into a more friendly and understandable conversation expression. It is said.
[0008]
[Means for Solving the Problems]
That is, the conversation expression generation device A1 according to the first aspect of the present invention generates a conversation expression based on monologue text composed of monologous text as shown in FIG. 1, and stores the monologue text. The monologue text acquisition means 1 that is acquired from the monologue text storage unit MTD, the preprocessing means 2 that divides the acquired monologue text into a single sentence format and generates one or more single sentence texts, and analyzes the sentence end expression of the generated single sentence text The sentence ending processing means 3 for associating the sentence ending expression with any one of a plurality of preset sentence ending expression patterns, and a plurality of comment texts associated with each sentence ending expression pattern and set as expressions to respond to them Single sentence associated with any one of the sentence end expression patterns from the comment storage unit CMD Comment selecting means 4 for selecting one comment text corresponding to the text, conversation expression generating means 5 for inserting the selected comment text after the single sentence text and generating a conversation text composed of the single sentence text and the comment text, and generation The conversation text output means 6 for outputting the conversation text is provided, and the above-described functions are achieved by operating the computer based on the program for the conversation expression generation device A1. The same is true in that each conversation expression generation device described below functions based on a dedicated program.
[0009]
Here, as described above, the monologous sentence is a sentence unilaterally described by a single writer, and the monologue text means text data composed of such a sentence. The single sentence text means a text composed of only one sentence ending with a punctuation mark. In addition, the monologue text storage unit MTD may store the monologue text as a database, or may store the input monologue text in a temporary memory.
[0010]
Such a conversation expression generation device A1 can automatically convert monologue text, which is generally difficult to understand, into conversation text composed of expressions in a conversation format that is easier to understand. Can reduce the burden on the person to understand the information. In particular, unlike conventional systems that use a conversation format, the appropriate conversation text is generated by focusing on the sentence end expression after monologizing the monologue text. It is completely different from the conventional one in that it can easily handle monologue sentences without restriction of newly created fields, not just using sentences as they are or realizing only conversation expressions in a defined field. And it has a wide range of applications.
[0011]
As shown in FIG. 2, the conversation expression generation device A2 according to the second aspect of the present invention further includes a keyword acquisition means 7 for acquiring a keyword input by the user in addition to the configuration of the conversation expression generation device A1. It has, Said Monologue text acquisition means 1 Said One or more corresponding to the keyword from the monologue text storage unit MTD Said Get the monologue text, Said The preprocessing means 2 is configured to generate a single sentence text for each of the acquired monologue texts.
[0012]
With such a configuration, when the user inputs a keyword of interest, the conversation text generated from the corresponding monologue text is output, so that the topic of interest can be further understood.
[0013]
In such conversation expression generation devices A1 and A2, Said The monologue text storage unit MTD receives the opinion text posted on the electronic bulletin board by user input. Said Is stored as monologue text, Said If the monologue text acquisition means 1 acquires the monologue text as the opinion sentence, it becomes easy to understand the opinions of various people.
[0014]
As shown in FIG. 3, the conversation expression generation device A3 according to the third aspect of the present invention includes, in addition to the configuration of the conversation expression generation device A1 or A2, each of two or more preset speaker agents. Against the output Said In addition to associating one speaker agent as the reader of the single sentence of the conversation text, Said The apparatus further includes speaker determining means 8 for performing processing for associating other speaker agents as readers of the comment text. In this conversation expression generation device A3, the keyword acquisition means 7 indicated by a broken line in FIG. 3 is not an essential function, and can be selected as an option to provide this function. Also, in each conversation expression generation device shown in FIGS. 4 to 7 described below, each means indicated by a broken line indicates an optional function in the device or program.
[0015]
In the case of such a conversation expression generation device A3, the conversation text that is a conversation form is composed of a single sentence text and a comment text, and these are distinguished as readers by different speaker agents. When displaying single sentence text and comment text on the display to be used, if two or more speaker agents associated with the single sentence text or comment text are displayed at the same time, two or more people can actually talk. Since it is possible to express the state as if the person is interacting with the user, the user who is the listener is very familiar and the degree of understanding can be improved.
[0016]
As shown in FIG. 4, the conversation expression generation device A4 according to the fourth aspect of the present invention, which further improves the effect of the conversation expression generation device A3, Said In addition to the configuration of the conversation expression generation device A3, Said It further comprises voice output means 9 for outputting a single sentence text or comment text corresponding to each speaker agent determined by the speaker determination means 8 with a different voice. In other words, by outputting the single sentence text and the comment text in different voices as if different speaker agents are speaking, the understanding of the listener is further improved.
[0017]
And as shown in FIG. 5, the conversation expression generation device A5 according to the fifth aspect of the present invention, which further improves the effect of the conversation expression generation device A4, Said In addition to the configuration of the conversation expression generation device A4, Said Corresponding to the voice of the single sentence text or comment text output by the voice output means 9, the animation processing means 10 further outputs to each speaker agent an animation operation that can move at least the mouth of the image of the speaker agent. It is provided. That is, in addition to voice data of simple text or comment text as if the speaker agent is actually speaking, by operating the image of the speaker agent, Mentioned above It is possible to advance the conversation in a more realistic manner than in the case of the conversation expression generation device A4.
[0018]
further, Said The conversation expression generation device A6 according to the sixth aspect of the present invention having an auxiliary function for any of the conversation expression generation devices A3, A4, A5, Said Each speaker agent is further provided with character data output means 11 for outputting the corresponding single sentence text or comment text as character data that can be displayed on the screen. With such a configuration, it is possible to display single text and comment text output as character data in association with each speaker agent and display it on the display, so even a deaf user can understand the content. It can be deepened, and by outputting together with audio data and animation images, a general user can easily grasp the contents by receiving information from the eyes and ears.
[0019]
Also, Said Although the case where the monologue text is accompanied by image data is also conceivable, the conversation expression generation device A7 according to the seventh aspect of the present invention in this case is described above. Said In addition to the configuration of each conversation expression generation device A1, A2, A3, A4, A5, A6, Said Image data processing means 12 for acquiring and outputting the image data from the monologue text storage unit MTD may be further provided. That is, the output image data can be used as an aid for understanding the conversation text.
[0020]
In particular, in any of the conversation expression generation devices A3, A4, A5, A6, Said As one of the speaker agents, set up a main caster agent to be a facilitator of conversation expression, Said When the speaker determining means 8 determines the main caster agent as the reader of the comment text, the conversation expression can proceed smoothly. And Said If the monologue text consists of a body part that is an essential part of the content and a title part that gives an overview, Said The speaker determining means 8 may be configured to determine the main caster agent as the reader of the title part. Also, Said If comment text indicating the start of the conversation expression is stored in the comment storage unit CMT, Said In the comment selection means 4, when there is no other single text before the single text, the comment text indicating the start is selected, Said By configuring the speaker selection means 8 so that the comment text combined with the title part is associated with the main caster agent, the flow of conversation can be made smooth. On the other hand, when one or more announcer agents are set as speaker agents different from the main caster agent, Said If the speaker determining means 8 determines the announcer agent as the reader of the text part, the division of roles with the main caster agent can be made clear.
[0021]
Explained above Said In order to make the flow of a plurality of conversation expressions based on a plurality of monologue texts smooth in any of the conversation expression generation devices A1 to A7, Said The comment selection means 4 recognizes the last single sentence text in one monologue text and associates it with the sentence end expression pattern in the last single sentence text to set the expression connected to the next monologue text and stores it in the comment storage section It is preferable that the comment text corresponding to the sentence end expression of the final single sentence text is selected from the text.
[0022]
In addition, sentence end expressions can be classified and patterned to some extent. That is, Said At the end of the sentence pattern, include at least a phenomenon description form that indicates the phenomenon and a hearing form that indicates the hearing, Said The comment text includes at least comment text corresponding to the question sentence format corresponding to the phenomenon description format and comment text corresponding to the expected sentence format corresponding to the hearing format, Said In the sentence end processing means 3, Said Corresponding to the sentence ending expression or the listening style of the sentence sentence end expression Said In the comment selection means 4, a mode in which one of the question text format or the predicted text format is selected as a preferred text form. In addition, the question sentence format and the expected sentence format Said Set multiple comment texts, Said If the comment selection means 4 selects any one comment text from the plurality of comment texts, a variation can be provided so that the conversation does not become monotonous.
[0023]
By the way, recently, an interactive broadcasting system for a community called a public opinion channel (hereinafter referred to as “POC”) is being developed. In this POC, community members post opinions on electronic bulletin boards directed to other members, and after the conversation generation processing is added to these opinions, the main caster agent who is a virtual speaker The opinion sentence is broadcast to community members in the form of an opinion introduction program based on the conversation of the announcer agent. Therefore, according to the present invention Said The conversation expression generation devices A1 to A7 have the monologue text storage unit for storing the monologue text input to the electronic bulletin board that can be used by a plurality of users input, and the comment storage unit. Applicable to interactive broadcasting systems such as POC that can broadcast conversational expression text generated based on the monologue text, and can generate conversation text based on the input monologue text and broadcast the conversation text Can play an extremely important role in opinion introduction broadcasting for communities such as POC.
[0024]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
[0025]
This embodiment is a conversation expression generating device applied to the above-mentioned public opinion channel (hereinafter referred to as “POC”) as shown in FIG. The conversation expression generation device A7 of the aspect is used. Hereinafter, the conversation expression generation device A7 is referred to as a POC caster A7. The POC provides a personal computer used by each user U who is a community member, a client computer CC such as a PDA or a mobile phone, and an electronic bulletin board that the user U can access from the client computer CC to the client computer CC and a client computer. It is basically composed of a POC server PS and POC caster A7 for storing the opinion sentence of the user U posted from the CC, and these client computers CC, POC server PS, and POC caster A7 can communicate bidirectionally through the Internet IN. It is connected to the. The POC caster A7 also has a function as a broadcasting client for broadcasting a conversational broadcast generated based on the opinion sentence posted from the user U to the client computer CC. The broadcast can be viewed on the screen of the client computer CC.
[0026]
First, the internal device configuration of each device will be described. The POC server PS is constituted by a general-purpose server computer and has a database server function and a Web server function. Among them, the Web server provides a home page and an electronic bulletin board that can be browsed from the client computer CC. In addition, the database server has a function as a monologue text storage unit MTD for storing the opinion sentence input to the electronic bulletin board. On the other hand, the POC caster A7 is constructed by a general server computer or personal computer. As shown in FIG. 9, the CPU 101, the internal memory 102, a storage device 103 such as a hard disk, and an input device such as a keyboard and a mouse. 104, an output device 105 such as a display and a speaker, various communication interfaces 106, and the like are included as internal devices. Note that the database device 107 may be further included as an internal device or an external device. Then, the program recorded in the storage device 103 is read into the internal memory 102 in accordance with an instruction from the CPU 101, necessary data is read out from the database device 107 as appropriate, and information is transmitted to the POC server PS and client computer CC via the communication interface 106. The POC caster A7 operates by performing communication. Note that when the POC caster A7 requires information input or screen display output, the input device 104 or the output device 105 is used as appropriate. In this embodiment, the POC server PS and the POC caster A7 are shown as separate computers capable of two-way communication through the Internet IN. However, these may be connected by a dedicated communication line or a single computer. It may be realized by a computer. Further, as described above, the client computer CC is composed of a general personal computer, PDA, mobile phone, or the like. Here, at least a connection function to the Internet IN, a character / image input / output function, a display, etc. And an audio output function such as a speaker.
[0027]
Next, the function of the POC caster A7 will be described. In the POC caster A7, each internal device and external device cooperate with each other in accordance with instructions from the CPU 101 based on the conversation expression generation program. As shown in FIG. Means 1, preprocessing means 2, sentence end processing means 3, comment selection means 4, conversation expression generation means 5, conversation text output means 6, keyword acquisition means 7, speaker determination means 8, voice output means 9, animation processing means 10 , Function as character data output means 11 and image data processing means 12. By installing a program for operating these means in the computer, the computer functions as the POC caster A7. This program is implemented by, for example, reading a program recorded on a recording medium such as a CD-ROM into a computer or downloading it to the computer through the Internet IN or the like. In the present embodiment, in particular, a TSS system (manufactured by Toshiba Corporation) is used for voice synthesis of agents (main caster agent MA and announcer agent AA) that are virtual speakers who conduct a conversation for introducing an opinion. Uses the photo face character creation system (Sharp Co., Ltd.) incorporated in the POC caster A7, but other products having functions equivalent to these can also be used.
[0028]
The database 107 functions as a comment storage unit CMD. Here, an example of the stored comment is shown in FIG. In this example, the sentence end expression patterns of opinion sentences posted by the user U are roughly divided into two types, one of which is further classified into three types, and a plurality of comments corresponding to each of them are prepared. More specifically, as shown in the left column of the figure, sentence ending expressions are broadly divided into a phenomenon description form that “states a phenomenon” and a “hearing form” that indicates that it is a hearing. Furthermore, the phenomenon description forms are “There is a form” that states the phenomenon, “Aspect form that states the phenomenon” (current / current form) ”,“ Aspect word that states the phenomenon “is” (Past / Past Progressive Type) ”. Examples of these sentence end expressions include those shown in the column of FIG. In other words, first of all, there are “~”, “has ~”, “has ~”, and “~” in the expression examples of “There is a form that describes the phenomenon”. Etc. Examples of the expression “the aspect word“ is ”that describes the phenomenon” (current / current progression) include “is”, “is”, “has been popular”. Can be mentioned. Examples of the expression “Ate” (representing the phenomenon) (past / progressive) are “it was” and “it was”. Examples of “hearing form” include “It seems to be ~”, “It seems to be ~”, “It is called ~”, “It is called ~”, and the like. Such an expression example is based on the fact that, as a result of analyzing the aspect of the sentence ending when introducing things, it has been found that the sentence is classified into a total of four types of sentence ending expression patterns as described above. The comment text to be inserted after each sentence end expression pattern is as shown in the right column of FIG. In other words, the “detailed question (current content)” and “detailed question (current content)” states “there are” forms that state the phenomenon and the “aspect” form that states the phenomenon (current / current progression). Comment texts such as "What is it?", "Tell me more", "What is it?", "What is it?" Also, in the “Aspect Word“ I ”(Past / Past Progressive)” that states the phenomenon, “How was it?”, “How was it?” As “Detailed Question (Past Content)” The comment text such as “.” Is associated. In addition, comment text such as “what is it?”, “What is it?”, “What is it?”, Etc., is associated with “hearing form” as “detailed prediction sentence”. Note that the expression examples and comment text examples described above are merely examples, and may include other than these. Although not shown in the figure, each comment text pattern and each comment text is given an appropriate identifier so as to be distinguished from other comment texts.
[0029]
However, in the database 107 functioning as the comment storage unit CMD, in addition to the above-described comment text, for example, a comment text such as “Yes” or “Is it?” That is simply an expression of matching, and an agent voices. Comment text and conversation such as “What do you think?”, Which is inserted between the opinion sentence that is virtually read out by the output and connected to the topic or calling other users One or more comment texts such as “Now it is a topic of ~” representing the start are also stored.
[0030]
In addition, the opinion sentence that the user U inputs at his / her client computer CC and submits by transmission is stored in the POC server PS. Each of these opinion sentences is a monologue composed of monologous sentences described by the user U himself / herself. Text. FIG. 11 shows an example of the opinion sentence. As shown in the figure, the opinion sentence OPT is composed of a title part OPH described in the “Title” field and a body part OPM described in the “Text” field. Is described. Each opinion sentence OPT is managed by a unique identifier that is distinguished from other opinion sentences OPT. In the present embodiment, the opinion sentence is converted into the XML (extensible Markup Language) format in the POC server PS, but may be in other formats. Furthermore, each opinion sentence may be accompanied by a related image OPI composed of a moving image or a still image.
[0031]
Further, it is assumed that two types of speaker agents are set in advance: a main caster agent and an announcer agent. That is, it is assumed that a facial character image created by the photoface character creation system and a synthesized voice created by the TSS system are set in association with each of the main caster agent and the announcer agent. Here, the announcer agent is set as a speaker who reads out the body part OPM of the original opinion sentence out of the conversation text created based on the opinion sentence. On the other hand, it is assumed that the main caster agent is set to play the title part OPH and the comment text.
[0032]
Hereinafter, an operation example of the POC caster A7 will be described using the example of the opinion sentence shown in FIG. 11, the example of the conversation text shown in FIG. 12, the example of the flowchart shown in FIG. 13, the example of the screen shown in FIG.
[0033]
First, as a premise, for example, it is assumed that many opinion sentences from the user U as shown in FIG. 11 are stored in the POC server PS as monolog text. In other words, each user U posts his / her opinion using the electronic bulletin board provided by the POC server PS. Further, the POC server PS or the POC caster A7 transmits and displays a screen displaying a keyword input field to the client computer CC of the user U who wants to view the introduction of the opinion sentence. It is assumed that a certain keyword is input to and transmitted from the client computer CC to the POC caster A7.
[0034]
When the POC caster A7 acquires the keyword transmitted from the client computer CC (FIG. 13; step S1), the POC caster A7 searches the POC server PS and searches for an opinion sentence (monolog text) OPT related to the acquired keyword (step S2). ). In this search, for example, an appropriate method such as a search of only the title part OPH, a full-text search of the title part OPH and the text part OPM can be employed. When one or more opinion sentences OPT can be acquired from the POC server PS (step S2a; Yes), the process proceeds to the next. Here, when there are a plurality of acquired opinion sentences OPT, they are arranged in an appropriate order such as an ascending or descending order of identifiers, or a date order. On the other hand, when there is no opinion sentence OPT corresponding to the acquired keyword (step S2a; No), information to that effect is transmitted to the client computer CC (step S2b).
[0035]
Next, it is determined whether or not all of the opinion sentences OPT have been introduced. If the introduction has been completed (step S3; No), a single sentence text in which one opinion sentence OPT is delimited for each period "." (Step S4). Then, the sentence end expression of each single sentence text is analyzed (step S5), the comment text corresponding to the sentence end expression pattern corresponding to the sentence end expression of each single sentence text is extracted from the database 107 (step S6), and the extracted comment text is extracted. The conversation text CVT is generated by inserting it after each single sentence text (step S7). Here, as an example, if the acquired opinion sentence OPT is as shown in FIG. 11, the generated conversation text CVT is as shown in FIG. That is, the body part OPM of the opinion sentence OPT shown in FIG. 11 is divided into four single sentence texts. First, the title part OPH is allocated to the main caster agent MCA. Here, since there is no preceding single sentence text in the opinion sentence OPT before the title part OPH, the comment text indicating the start of the conversation and the description of the title part OPH are synthesized, and “First of all, walking. "This is a topic." In this way, whether or not “first time ...” can be changed as appropriate according to the order of introduction of the opinion sentences. For example, comments such as “Next is ...” or “Last is ...” You can use text. Next, the first sentence of the body part OPM is allocated to the announcer agent ANA, but the sentence at the end of the first minute corresponds to the “hearing form” of “… Yes.” The comment text is assigned to the main caster agent MCA. Similarly, the second sentence and the third sentence are allocated to the announcer agent ANA, and comment texts “Tell me more” and “Yes” corresponding to the sentence end expressions of the second sentence and the third sentence are main. Allocated to the caster agent MCA. Furthermore, the fourth sentence is also allocated to the announcer agent ANA, but since the fourth sentence is a single sentence text at the end of the opinion sentence OPT, it is an expression that is called to another user U, and is sent to the next opinion sentence OPT. The comment text “How about you,” which will be connected, is assigned to the main caster agent MCA.
[0036]
Returning to the flowchart shown in FIG. 13, the description of the title part OPH among the generated conversation text CVT (step S8; Yes) gives the main caster agent MCA the comment text utterance operation, while the announcer A break operation is given to the agent ANA (step S8a). If it is not the title part OPH, that is, the text part OPM (step S8; No), it is determined whether or not the sentence end expression is an expression that states introduction, and if not (step S8b; No), The announcer agent ANA is given a reading operation of a single text, while the main caster agent MCA is given a resting operation (step S8c). On the other hand, if the sentence ending expression is an expression that describes the introduction (step S8; Yes), the main caster agent MCA is given a comment text reading operation, while the announcer agent ANA is given a rest operation (step S8a). Then, when an action is given to the main caster agent MCA and the announcer agent ANA, voices are synthesized when the corresponding animation is generated. Here, the animation operation includes an operation in which at least the main caster agent MCA and the announcer agent ANA can move their mouths. However, as will be described later, if the attached image is attached to the opinion sentence OPT, There are variations, such as making the finger point and move. Further, as described above, if there is an attached image OPI in the opinion sentence OPT (step S10; Yes), the attached image is transmitted, for example, by placing it in the center of the transmission image to be displayed on the client computer CC. Combine the attached image with the image. After that, or when there is no attached image (step S10; No), the text data of the conversation text CVT is synthesized in, for example, the lower column of the transmitted image (step S11), and all data is transmitted in a format that can be viewed on the client computer CC. (Step S12). The images displayed on the display of the client computer CC as a result of the transmission are as shown in FIGS. 14 to 22, for example. The audio of the main caster agent MCA and the announcer agent ANA is transmitted from the speaker of the client computer CC. Is output.
[0037]
The screen displayed on the display of the client computer CC and the sound output from the speaker will be described. In FIGS. 14 to 22, an attached image is displayed at the center of the screen, and an image of the main caster agent MCA is displayed on the left side. The image of the announcer agent ANA is arranged on the right side. Then, in accordance with the conversation text CVT shown in FIG. 12, the action of each agent and the voice output for reading out the comment or the opinion sentence are sequentially displayed, and the character data associated with this voice output is displayed (the lower column of the screen). First, as shown in FIG. 14, the main caster agent MCA performs an operation of reading out the comment “First of all ...” which is a combination of the title part OPH of the opinion sentence and the comment text indicating the start of the conversation, Output and display the character data of this comment in the lower column of the screen. Next, as shown in FIG. 15, the announcer agent ANA reads out the first sentence “Ise Shimano…” in the body part OPM, outputs the voice, and displays the character data of this comment in the lower column of the screen. Is displayed. At this time, the announcer agent ANA is caused to perform an animation operation indicating the attached image OPI. Next, as shown in FIG. 16, the speaker switches to the main caster agent MCA to read out the comment “What is it?” And outputs the voice, and the comment is displayed in the lower column of the screen. Display character data. Furthermore, as shown in FIG. 17, the speaker replaces the announcer agent ANA and reads out the second sentence “Large-scale facility sightseeing…” in the text part OPM, and outputs the voice, The character data of this comment is displayed in the column. Subsequently, as shown in FIG. 18, the main caster agent MCA performs an operation of reading out the comment “Tell me more”, outputs the sound, and displays the character data of this comment in the lower column of the screen. Next, as shown in FIG. 19, the speaker changes to the announcer agent ANA, reads out the third sentence “Beautiful scenery and…” of the text part OPM, outputs the voice, and displays it in the lower column of the screen. Displays the character data of this comment. On the other hand, as shown in FIG. 20, the main caster agent MCA performs an operation of reading out the comment “Yes.”, Outputs the voice, and displays the character data of this comment in the lower column of the screen. Then, as shown in FIG. 21, the announcer agent ANA reads out the final sentence “Would you like to join us?” In the body part OPM, outputs the sound, and displays the comment text in the lower column of the screen. Display data. Finally, as shown in FIG. 22, the main caster agent MCA reads out the comment “How about you, everyone”, outputs the sound, and displays the character data of this comment in the lower column of the screen. This leads to the introduction of the next opinion sentence OPT.
[0038]
That is, when the above steps are completed, the process returns to step S3, and the next opinion sentence OPT is processed. When the introduction of all the opinion sentences OPT ends (step S3; Yes), the processing for the originally acquired rank keyword is ended. Note that steps S9 to S11 are not necessarily in this order, and the order may be appropriately changed.
[0039]
As described above, for the opinion sentence posted by the user U, the intention of the target opinion sentence is inferred from the sentence end expression, which is a surface clue of the opinion sentence, and the comment text is inserted or added and synthesized. Therefore, the other user U can view the opinion sentence in a conversation format. Therefore, even if the original information presentation is an opinion sentence that is a monologous sentence, information can be provided in a manner that is more familiar to the user U who views it and that reduces the burden on understanding.
[0040]
The present invention is not limited to the embodiment described above. For example, it is possible to increase or decrease the sentence end expression patterns of the divided single sentence text, and to appropriately increase or decrease the expression examples corresponding to each pattern and the number of corresponding comment texts. In addition, since the text posted on the electronic bulletin board is not a formal text, the title section and the text section may be connected as one text in the posted opinion text. It is also possible to perform processing such as normalizing a sentence that begins with a case particle such as “O”, “GA”, “NO”, or a symbol such as “...”. In addition, the specific configuration of each part is not limited to the above embodiment, and various modifications can be made without departing from the spirit of the present invention. Further, the generated conversation text is not limited to two, but can be a conversation sentence of three or more parties. Furthermore, the present invention can be applied to fields or systems other than POC.
[0041]
【The invention's effect】
According to the present invention, as described in detail above, a monologue text, which is a sentence described by one person, is processed into a single sentence, and the conversation format that can reduce the burden on the understanding required by the listener or the reader. It can be converted. In other words, by patterning the sentence end expression, which is the surface clue of the monologue text, guessing the intended purpose of the monologue text and inserting an appropriate comment at the end of the single sentence, it is possible to add to the topic content or field. It is possible to generate a conversation expression without limitation. Therefore, by applying the present invention, it is greatly useful for the management of opinion introduction programs in a community where unspecified topics such as POC appear, and for research on conversation.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a functional configuration of a conversation expression generation device corresponding to claim 1 of the present invention;
FIG. 2 is a block diagram showing a functional configuration of a conversation expression generation device corresponding to claim 2 of the present invention;
FIG. 3 is a block diagram showing a functional configuration of a conversation expression generation device corresponding to claim 4 of the present invention;
FIG. 4 is a block diagram showing a functional configuration of a conversation expression generation device corresponding to claim 5 of the present invention;
FIG. 5 is a block diagram showing a functional configuration of a conversation expression generation device corresponding to claim 6 of the present invention;
FIG. 6 is a block diagram showing a functional configuration of a conversation expression generation device corresponding to claim 7 of the present invention;
FIG. 7 is a block diagram showing a functional configuration of a conversation expression generation device corresponding to claim 8 of the present invention;
FIG. 8 is an overview diagram showing a POC system to which an embodiment of the present invention is applied.
FIG. 9 is a schematic internal device configuration diagram of the POC caster of the embodiment.
FIG. 10 is a diagram showing an example of internal data of a comment storage unit applied to the embodiment.
FIG. 11 is a diagram showing an example of an opinion sentence applied to the embodiment.
FIG. 12 is a view showing an example of a conversation text generated in the embodiment.
FIG. 13 is a flowchart schematically showing the operation of the embodiment.
FIG. 14 is a diagram showing an example of a screen displayed on the client computer in the embodiment.
FIG. 15 is a view showing an example of a screen displayed on the client computer in the embodiment.
FIG. 16 is a view showing an example of a screen displayed on the client computer in the embodiment.
FIG. 17 is a view showing an example of a screen displayed on the client computer in the embodiment.
FIG. 18 is a view showing an example of a screen displayed on the client computer in the embodiment.
FIG. 19 is a view showing an example of a screen displayed on the client computer in the embodiment.
FIG. 20 is a view showing an example of a screen displayed on the client computer in the embodiment.
FIG. 21 is a view showing an example of a screen displayed on the client computer in the embodiment.
FIG. 22 is a view showing an example of a screen displayed on the client computer in the embodiment.
[Explanation of symbols]
1 ... Monologue text acquisition means
2 ... Pretreatment means
3 ... sentence end processing means
4 ... Comment selection means
5 ... Conversation expression generating means
6 ... Conversation text output means
7 ... Keyword acquisition means
8 ... Speaker decision means
9. Sound output means
10: Animation processing means
11: Character data output means
12. Image data processing means
A1, A2, A3, A4, A5, A6, A7 ... conversation expression generating device
CMD ... Comment storage
MTD: Monologue text storage

Claims (32)

Based on monologue text composed of monologue text, a conversation expression is generated. Monologue text acquisition means for acquiring from the monologue text storage unit storing monologue text, and the obtained monologue text is divided into a single sentence format. Preprocessing means for generating the above single sentence text, sentence end processing means for analyzing the sentence end expression of the generated single sentence text and associating the sentence end expression with any one of a plurality of preset sentence end expression patterns, and each sentence end One comment text corresponding to the single sentence text associated with any one sentence end expression pattern from the comment storage unit that stores a plurality of comment texts set as expressions responding to the expression patterns Comment selecting means to select and selected after the simple text A conversation expression generating means for generating a conversation text by inserting the instrument text consisting sentence text and comment text conversation expression generating apparatus characterized by comprising; and a conversation text output means for outputting the generated conversation text. Further comprising a keyword obtaining means for obtaining the input keyword by a user, the monologue text acquisition means, the monologue be from the text storage unit configured to acquire one or more of the monologue text corresponding to the keyword, the The conversation expression generation apparatus according to claim 1, wherein the preprocessing means generates a single sentence text for each of the acquired monologue texts. But the monologue text storing unit, which stores an opinion statement text posted on the electronic bulletin board by an input of the user as the monologue text, said monolog text acquisition means acquires the monologue text is this sentiments The conversation expression generation device according to claim 1 or 2. For each of the two or more preset speaker agent, together with the associated one of the speaker agent as reader's simple sentence text of the outputted the conversation text, the other speaker agent as readers of the comment text 4. The conversation expression generation device according to claim 1, further comprising speaker determination means for performing a process of associating. 5. The conversation expression generating apparatus according to claim 4, further comprising voice output means for outputting a single sentence text or comment text corresponding to each speaker agent determined by the speaker determination means by voice. In correspondence with the voice of the single sentence text or comment text output by the voice output means, animation processing means for adding an animation operation that can move at least the mouth of the image of the speaker agent to each speaker agent and outputting it is further provided. The conversation expression generation device according to claim 5. Wherein with each speaker agent, the corresponding simple sentence text or conversation representation generator of the character data output means further comprising to have claim 4, 5 or 6, wherein the outputting the comment text as screen printable character data. 5. The main caster agent that is a facilitator of conversation expression is set as one of the speaker agents, and the speaker determining means determines the main caster agent as a reader of comment text. The conversation expression generation device according to 5, 6 or 7 . The monologue text is composed of a body part which is an essential part of the content and a title part indicating an outline, and the speaker determining means determines the main caster agent as a reader of the title part. conversation representation generation apparatus is claimed in claim 8. Comment text indicating the start of conversation expression is stored in the comment storage unit, and the comment selection means selects the comment text indicating the start when there is no other single text before the single text, 10. The conversation expression generation device according to claim 9 , wherein the speaker selection means associates the comment text with the title part combined with the main caster agent . The monologue text is composed of a body part which is an essential part of the content and a title part indicating an outline, and one or more announcer agents are set as speaker agents different from the main caster agent. 11. The conversation expression generating device according to claim 8, wherein the speaker determining means determines the announcer agent as a reader of the body part . The image data processing means which acquires and outputs the said image data from the said monologue text storage part when the said monologue text accompanies image data, It further comprises, The conversation expression generation device according to 7, 8, 9, 10 or 11. The comment selecting means recognizes the last single sentence text in one monologue text and associates it with the sentence end expression pattern in the last single sentence text to set the expression to be connected to the next monologue text and stored in the comment storage unit The conversation expression according to claim 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12, wherein the comment text corresponding to the sentence end expression of the final single sentence text is selected from the text. Generator. The suffix patterns, with a hearsay form indicating that phenomenon is hearsay symptoms and narrative showing that make mention is included at least, a comment corresponding to the question sentence form the comment in the text, corresponding to the phenomenon narrative text and a comment text corresponding to the expected sentence format corresponding to hearsay format contains at least the end of the sentence processing means, together with associated to one of the simple sentence phenomenon narrative or hearsay form endnote representation of text, correspondingly in the comment selection means Te, and selects either one of the comment text in question form or expected sentence format claim 1, 2, 3, 4 , 12 or 13. The comment text in question format and expected sentence format, respectively Yes set by plurality, the comments selection means, according to claim 14, wherein it is intended to select any one of the comment text from the plurality of comment text Conversation expression generator. The monologue text storage unit that stores monologue text input to an electronic bulletin board that can be used by a plurality of users and the comment storage unit, and is generated based on the input monologue text An interactive broadcasting system capable of broadcasting a conversation expression text to be generated, used to generate a conversation text based on an input monologue text, and to output the conversation text so as to be broadcastable. 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15. A program for causing a computer to function as a conversation expression generation device that generates a conversation expression based on monolog text composed of monologous sentences by operating the computer, the computer storing the monolog text Monologue text acquisition means to acquire from the text storage, preprocessing means to divide the acquired monologue text into a single sentence format and generate one or more single sentence texts, analyze the sentence end expression of the generated single sentence text, and The sentence end processing means for associating with any one of a plurality of sentence end expression patterns set in advance, and a comment storage unit for storing a plurality of comment texts set as expressions corresponding to each sentence end expression pattern and responding thereto Any one sentence end expression pattern Comment selecting means for selecting one comment text corresponding to the associated single sentence text, and conversation expression generating means for inserting the selected comment text after the single sentence text and generating conversation text composed of the single sentence text and the comment text And a conversation expression generating program that functions as a conversation text output means for outputting the generated conversation text. Said computer to further function as a keyword acquisition means for acquiring a keyword inputted by a user, wherein in monologue text acquisition unit acquires one or more monologue text corresponding to the keyword from the monologue text storage unit, the front The conversation expression generation program according to claim 17, wherein the processing means generates a single sentence text for each of the acquired monologue texts. The monologue text storage unit, which stores an opinion statement text posted on the electronic bulletin board by an input of the user as the monologue text, in the monologue text acquisition means, so as to obtain a monologue text is this sentiments The conversation expression generation program according to claim 17 or 18. The computer, for each of the two or more preset speaker agent, together with the associated one of the speaker agent as reader's simple sentence text of the outputted the conversation text, the other as a reader of the comment text 20. The conversation expression generation program according to claim 17, 18 or 19, further functioning as speaker determining means for performing processing for associating a speaker agent. The computer, the corresponding simple sentence text or claim 20 conversation representation generator according to further features comments text as speech output means for audio output with different sound for each speaker agents determined in the speaker determining means. Said computer, in response to the sound of simple sentence text or comment text is output by the speech output unit, said adding animation operation move at least the mouth of each speaker agent of the speaker's agent image output animation The conversation expression generation program according to claim 21, which further functions as processing means. Said computer, said with each speaker agent, conversation representation of the corresponding simple sentence text or even claims 19, 21 or 22, wherein the functioning of the said comment text as character data output means for outputting a screen printable character data generation program. The main caster agent that is a facilitator of conversation expression is set as one of the speaker agents, and the speaker determining means determines the main caster agent as a reader of the comment text. The conversation expression generation program according to 20, 21, 22, or 23 . The monologue text is composed of a body part which is an essential part of the content and a title part indicating an outline, and the speaker determining means determines the main caster agent as a reader of the title part. The conversation expression generation program according to claim 24 . Wherein the comment storing portion, are stored comments text indicating the start of a conversation representation, the selected comment text indicating the start in front of the simple sentence text if no other sentence text in the comment selection means, 26. The conversation expression generating program according to claim 25 , wherein said speaker selecting means associates the comment text with a title part combined with a main caster agent . The monologue text is composed of a body part which is an essential part of the content and a title part indicating an outline, and one or more announcer agents are set as speaker agents different from the main caster agent. 27. The conversation expression generation program according to claim 24, 25, or 26, wherein the speaker determination means determines the announcer agent as a reader of a body part . When the monologue text is accompanied by image data, the computer is further caused to function as image data processing means for acquiring and outputting the image data from the monologue text storage unit. The conversation expression generation program according to 22, 23, 24, 25, 26 or 27 . In the comment selection unit recognizes the last sentence text in one monologue text, in association with the suffix patterns in the final sentence text is set as an expression for connecting to the next monologue text stored in the comment storing portion 29. The comment text corresponding to the sentence end expression of the final single sentence text is selected from the comment text. 29. The method according to claim 18, wherein the comment text is selected from the comment text. Conversation expression generation program. The suffix patterns, with a hearsay form indicating that phenomenon is hearsay symptoms and narrative showing that make mention is included at least, a comment corresponding to the question sentence form the comment in the text, corresponding to the phenomenon narrative comments and text contains at least corresponding to the expected sentence format corresponding to the text and hearsay form, together with the associated endnote representation of simple sentence text in any of the symptoms narrative or hearsay form in the end of the sentence processing means, the correspondingly The comment selection means selects either one of the question text format or the expected text format comment text, 18, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28. The conversation expression generation program according to 28 or 29. The comment text in question format and expected sentence format, respectively Yes set by a plurality, in co said instrument selection means, claim they are a plurality of comment text to select any one of the comment text 30 The described conversation expression generation program. Said monologue text storage unit for storing a monologue text entered into the electronic bulletin board available by a plurality of user inputs, and a the comment storing portion, and, generated based on the inputted monologue text 18. The present invention is applied to an interactive broadcasting system capable of broadcasting a conversation expression text, and is used to generate a conversation text based on an inputted monologue text and to output the conversation text so as to be broadcastable. The conversation expression generation program according to 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 or 31.

Images