KR102183935B1 - Conversation making device using personality information - Google Patents

Abstract

The present invention relates to a conversation generator using personality information, and is connected to a character-specific word use characteristic rule DB that matches and stores at least one personality-specific word with respect to at least one personality information, and the personality-specific word use characteristic rule DB. , A word selector for selecting a word for each personality using a word for each personality information stored in the personality rule DB, according to the input previous conversation content and the input personality information, at least one for at least one personality information Sentence generation characteristic rule DB for each personality, which matches and stores the sentence generation characteristic rule for each personality, and a sentence by combining the words selected from the word selector according to the input personality information and connected to the sentence generation characteristic rule DB for each personality. A conversation to be provided after the previous conversation content is automatically generated based on the input personality information. For this reason, it is possible to automatically generate a conversation having a specific personality as well as a counselor's conversational body using a word database for each personality, and thus there is an effect that a conversation having a specific dialogue can be utilized in various fields.

Description

Conversation generator using personality information {CONVERSATION MAKING DEVICE USING PERSONALITY INFORMATION}

The present invention relates to a conversation generator using personality information, and more particularly, to a technology for automatically generating conversation sentences by reflecting personality information.

An interactive automatic chat service (chatbot, chatbot) that provides answers in sentences to simple keyword input or simple query sentence input has been disclosed. However, since such chatbots are characterized by providing predetermined answers to simple keywords, there is a limitation in providing conversations in addition to answers to questions.

On the other hand, in addition to the simple question and answer, the conversation created by the conversation generator that creates the conversation content has a unified conversational object, so there is a limitation in the conversational object to be applied to other conversation generation services other than the automatic agent conversation service.

Accordingly, there is a need to develop a conversation generator capable of automatically generating a conversation having a plurality of conversation objects.

Republic of Korea Patent Publication No. 10-0546732 (automatic question/correction indexing method based on language analysis and its question-and-answer method and system)

The technical problem to be achieved by the present invention is to automatically generate and provide a conversation having multiple conversations using a word database for each character in the conversation generator automatically generating conversations.

A conversation generator using personality information according to an embodiment of the present invention is configured to match and store at least one personality word with respect to at least one personality information. A word selector that selects a word for each personality using the words for each personality information stored in the personality word usage characteristic rule DB, according to the connected, inputted previous conversation content and input personality information, at least for at least one personality information A character-specific sentence generation characteristic rule DB that matches and stores one character-specific sentence generation characteristic rule, and a word selected from the word selector is combined according to the input personality information and connected to the character-specific sentence generation characteristic rule DB. And a sentence generator for generating a sentence, and automatically generates a conversation to be provided after the previous conversation content based on the input personality information.

The word selector selects a word to be used for the next conversation according to the previous conversation, but searches for the input personality information in the word use characteristic rule DB for each personality and selects a word for the input personality information. A processor, and a communication module that transmits the word selected by the word selection processing processor to the outside, or receives the previous conversation content or the personality information and transmits it to the word selection processing processor, wherein the sentence generator comprises: the input Sentence generation processing processor for generating sentences using the words selected by the word selection processing processor by searching the personality information in the sentence generation characteristic rule DB for each personality, and transmitting the sentences generated by the sentence generation processing processor to the outside Or a communication module for receiving the selected word from a communication module of the word selector and transmitting the selected word to the sentence generating processor, or transmitting the sentence generated by the sentence generating processor to the outside.

The word selector includes a storage device for storing a word selected by the word selection processing processor, an output device outputting the word selected by the word selection processor, and the previous conversation content or the personality information, and processing the word selection Further comprising an input/output device including an input device transmitted to the processor, wherein the sentence generator comprises: a storage device for storing a sentence generated by the sentence generating processor, and outputting a sentence generated by the sentence generating processor And an input/output device including an output device and an input device that receives the personality information and transmits the received personality information to the sentence generation processing processor.

The personality-specific word use characteristic rule DB is characterized in that the personality information and words for each personality information are additionally input and updated.

The personality-specific word use characteristic rule DB is characterized in that the frequency of the personality information-specific words referenced by the word selector is matched with the personality information-specific words and stored.

A sentence combination characteristic rule DB for each personality that matches and stores at least one sentence combination characteristic rule for each personality information, and is connected to the sentence combination characteristic rule DB for each personality, and according to the input personality information, It characterized in that it further comprises a conversation generator for generating a conversation by combining sentences using the sentence combination characteristic rule for each personality information stored in the character-specific sentence combination characteristic rule DB.

The conversation generator comprises a conversation generation processing processor for generating a conversation for the input personality information by searching for the input personality information in the sentence combination characteristic rule DB for each personality, and a conversation generated by the conversation generation processing processor. It characterized in that it comprises a communication module for transmitting to the outside or receiving the personality information to the conversation generation processing processor.

The conversation generator receives a storage device for storing a conversation generated by the conversation generation processing processor, an output device outputting the conversation generated by the conversation generation processing processor, and the personality information, and transfers the received personality information to the conversation generation processing processor. It characterized in that it further comprises an input/output device including an input device to perform.

The storage device of the word selector and the storage device of the sentence generator match personality information stored in the character-specific word usage characteristic rule DB and similar personality information with respect to personality information stored in the character-specific sentence generation characteristic rule DB It is characterized in that to store.

The word selection processing processor and the sentence generation processing processor, when the input personality information does not exist in the character-specific word use characteristic rule DB or the personality-specific sentence generation characteristic rule DB, similar personality information stored in the storage device Searching for, extracting the replacement personality information, and searching for the extracted alternative personality information in the character-specific word use characteristic rule DB or the personality-specific sentence generation characteristic rule DB to select a word or generate a sentence.

According to this feature, the conversation generator using personality information according to an embodiment of the present invention can automatically generate not only the agent conversation body but also the conversation having a specific personality using the personality word database, so that a conversation having a specific conversation body There is an effect that can be used in various fields.

1 is a block diagram showing a schematic structure of a conversation generator using personality information according to an embodiment of the present invention.
2 is a block diagram schematically showing the structure of a word selector among dialog generators using personality information according to an embodiment of the present invention.
3 is a block diagram schematically showing the structure of a sentence generator among conversation generators using personality information according to an embodiment of the present invention.
4 is a block diagram schematically showing the structure of a conversation generator among conversation generators using personality information according to an embodiment of the present invention.
5 is a diagram showing a similar personality information matching structure stored in a storage device of the conversation generator among conversation generators using personality information according to an embodiment of the present invention.
6 is a block diagram showing a schematic structure of a conversation generator using personality information according to another embodiment of the present invention.
7 is a block diagram showing a schematic structure of a conversation generator using personality information according to another embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art may easily implement the present invention. However, the present invention may be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and similar reference numerals are assigned to similar parts throughout the specification.

1 is a block diagram showing a schematic structure of a conversation generator using personality information according to an embodiment of the present invention, and FIG. 2 is a structure of a word selector among conversation generators using personality information according to an embodiment of the present invention. 3 is a block diagram schematically showing the structure of a sentence generator among dialog generators using personality information according to an embodiment of the present invention, and FIG. 4 is a block diagram showing personality information according to an embodiment of the present invention. Is a block diagram schematically showing the structure of a conversation generator among conversation generators using, and FIG. 5 is a similar personality information matching stored in a storage device of the conversation generator among conversation generators using personality information according to an embodiment of the present invention. Fig. 6 is a block diagram showing a schematic structure of a dialog generator using personality information according to another embodiment of the present invention, and Fig. 7 is a block diagram showing a schematic structure of a conversation generator using personality information according to another embodiment of the present invention. It is a block diagram showing the schematic structure of a dialog generator.

In the present embodiment with reference to FIGS. 1 to 7, the conversation generator S using personality information uses the personality information 10 and the previous conversation 20 and refers to the character-specific word use characteristic rule DB 110. A word selector 100 for selecting a word, and a sentence generator 200 for generating a sentence by using the word and personality information 10 selected by the word selector 100 and referring to the sentence generation characteristic rule DB 210 for each personality , And a dialogue generator 300 that uses the sentence and personality information 10 generated by the sentence generator 200 and generates a dialogue by referring to the sentence combination characteristic rule DB 310 for each personality.

In this case, the word selector 100 may include a word selection processor 101, a communication module 102, a storage device 103, and an input/output device 104. The word selection processing processor 101 is a processor that selects a word for the personality information 10 based on the personality information 10, the previous conversation 20, and the character-specific word use characteristic rule DB 110 to be.

In one example, the previous conversation 20 received by the word selection processing processor 101 may be received from an external server through the communication module 102, and the subject inputting the previous conversation 20 through the external server May be an external server, for example, a game user account that wants to access a game server and perform a conversation with a non-player character (NPC) in the game.

Alternatively, the previous conversation 20 received by the word selection processor 101 may be voice information of a customer input through an input pad of a kiosk or a microphone.

The word selection processing processor 101 selects a word to be used for the next conversation according to the received previous conversation 20, but searches the generated word in the character-specific word use characteristic rule DB 110 and selects it as a character-specific word.

In one example, when the content of the previous conversation 20 input to the word selector 100 is'Did you eat?' and the personality information 10 is'positive personality', the word selection processing processor 101 responds to the meal question. The word'yes' or'no', which is a word to be used in the answer conversation, is extracted, and any one of the extracted words,'yes' and'no', and'positive personality' are selected from the character-specific word use characteristic rule DB 110 Explore.

At this time, the character-specific word use characteristic rule DB 110 may match and provide words for personality information as shown in Table 1 below.

Personality information word 1st word by personality 2nd word by personality interjection positive Yeah Yeah of course haha no Still I'm going to eat it later Oh negative Yeah Whatever the case, what do you know. Why ate it! huh! no It's still! Sheesh!

Personality information included in the character-specific word use characteristic rule DB 110 may further include optimistic, pessimistic, extroverted, introspective, emotional, depressed personality, and capricious personality in addition to positive and negative personality, and may further include character-specific word use characteristics rule DB The number and types of personality information included in (110) are not limited.

As for the words matching the personality information included in the character-specific word use characteristic rule DB 110, as shown in Table 1 above, all words for conversation generation are matched and stored for one personality information, and all words are stored. The first word and the second word, which are words for each personality, are matched and stored. At this time, the first word that the positive personality information has for the word'yes' is'yes', and the first word that the negative personality information has for the word'yes' is'What will I know no matter what happens?' Is saved. In addition, the second word that the positive personality information has for the word'no' is stored as'I want to eat it sometime', and the second word that the negative personality information has for the word'no' is'still. It is saved as !'.

And, as described in Table 1 above, the character-specific word use characteristic rule DB 110 matches and stores the exclamation point for various personality information, and the exclamation point is stored as'haha' or'ahh' for positive personality information. Then, the words of'yes' or'no' matched with the positive personality information may be matched and stored, respectively. On the other hand, the exclamation for the negative personality information may be stored as'Hung!' or'Chet!', and the words of'Yes' or'No' matched with the negative personality information may be matched and stored, respectively.

In one example, the character-specific word use characteristic rule DB 110 may be updated by additionally receiving personality information and words therefor, and may also be updated according to an input signal for modifying stored word information for personality information. .

In addition, the character-specific word use characteristic rule DB 110 may additionally receive a word for generating a conversation, and receive and update a character-specific word for the added new word to match each personality information.

When the character-specific word use characteristic rule DB 110 includes a plurality of personality words for a specific word matched with one personality information as a first word and a second word, the plurality of words will be stored in a specific order. It may be stored according to an input signal that controls words that are frequently used for sentence generation to be located in an upper order.

As such, the character-specific word use characteristic rule DB 110 is a database referred to by the word selection processing processor 101 for word selection for personality information, and the character-specific word use characteristic rule DB 110 is a word selection processing processor ( The word information is extracted and provided to the word selection processing processor 101 according to a request for a word for each personality information to be referred to in 101).

Alternatively, in one example, the character-specific word use characteristic rule DB 110 matches and stores the number of times the word selection processing processor 101 refers to the word information for each personality information, so that the character-specific word use characteristic rule DB 110 When choosing a word, you can refer to the frequency of reference and let you choose a word.

In this way, the character-specific word use characteristic rule DB 110 can provide the words stored for various personality information to the word selection processing processor 101, so that the word selection processing processor 101 can be used for a conversation that follows the previous conversation. In extracting words, it is possible to extract frequently used words from specific personality information.

As shown in FIG. 1, the character-specific word use rule DB 110 may be formed as a database implemented separately from the word selector 100, but in another embodiment, the word selector 100 is provided together Thus, according to the request of the word selector 100, words for each personality information may be provided.

Alternatively, in another example, the character-specific word use characteristic rule DB 110 may be implemented as a server separate from the word selector 100, and the word generator 100 is the character-specific word use characteristic rule DB 110 A word for each personality information may be requested by accessing the formed server, and accordingly, the word selector 100 may receive a word for each personality information from the server, but is not limited thereto.

At this time, in an example in which the character-specific word use characteristic rule DB 110 is implemented as a server, the character-specific word use characteristic rule DB 110 transmits and receives signals and data through wired communication or wireless communication with the word selector 100 It may include a wired communication module or a wireless communication module.

Continuing to describe the word selection processing processor 101, the word selection processing processor 101 includes the personality information 10 and the previous conversation 20 based on the word information received from the character-specific word use characteristic rule DB 110. ), a specific word is selected and transmitted to the sentence generator 200.

In one example, the word selection processor 101 may transmit the selected word to the sentence generator 200 through the communication module 102. In this case, the communication module 102 may be a communication module that performs wired communication or wireless communication with the sentence generator 200. The communication module 102 may communicate with the character-specific word use characteristic rule DB 110, or may access the character-specific word use characteristic rule DB 110 formed as a server.

In addition, the word selection processing processor 101 may store the selected word in the storage device 103. In this case, the storage device 103 may be a volatile memory or a nonvolatile memory that stores the word selected by the word selection processor 101, and the selected word by receiving personality information 10 from the word selection processor 101 together Can be saved by matching to.

In addition, the word selection processing processor 101 may output the selected word to an output device of the input/output device 104. The input/output device 104 receives the personality information 10 and the previous conversation 20 and transmits it to the word selection processor 101, or a signal for controlling to update the stored data of the character-specific word usage characteristic rule DB 110 Also includes an input device that receives input.

In one example, the input device may be a mouse, keyboard, or microphone, and the output device may be a monitor or speaker, but is not limited thereto.

The input device of the input/output device 104 may directly receive any one of an optimistic personality, a pessimistic personality, an extrovert personality, an introspective personality, an emotional personality, a depressed personality, and a capricious personality as the personality information 10, or can be directly input to the output device. It may be input as a selection signal among a plurality of output personality information.

In one example, the plurality of personality information output from the output device of the input/output device 140 may be personality information defined in the character-specific word usage characteristic rule DB 110, as another example, the storage device 103 It can be the personality information you are storing in.

Again, referring to FIG. 1, when the sentence generator 200 of the dialogue generator S using personality information is described, the sentence generator 200 is connected to the sentence generation characteristic rule DB 210 for each personality, and a word selector ( As an apparatus for automatically generating a sentence by receiving word information selected for personality information from 100), as shown in FIG. 3, a sentence generation processing processor 201, a communication module 202, a storage device 203, and an input/output device ( 204).

The sentence generation processing processor 201 of the sentence generator 200 automatically generates sentences for each personality by referring to the personality information 10, the selected word received from the word selector 100, and the sentence generation characteristic rule DB 210 for each personality. It is a processor that generates by.

The sentence generation characteristic rule DB 210 for each personality referenced by the sentence generation processing processor 201 may be provided by matching sentence generation rules for personality information as shown in Table 2 below.

Personality information Characteristic Sentence Generation Trait Rules positive Kind and long sentence negative Short answer

Personality information included in the character-specific sentence generation characteristic rule DB 210 may further include optimistic, pessimistic, extroverted, introspective, emotional, depressed personality, and capricious personality in addition to positive and negative personality, and may further include character-specific sentence generation characteristic rule DB The number and types of personality information included in (210) are not limited.

The character-specific sentence generation characteristic rule DB 210 may include a plurality of character-specific sentence generation characteristic rules for one personality information, and may be updated to be added or modified according to an input signal. When the character-specific sentence generation characteristic rule DB 210 includes a plurality of character-specific character-specific character-specific rules for one personality information, the frequency referenced by the sentence generation processing processor 201 may be matched to each rule and stored. , The sentence generation processing processor 201 may extract a sentence generation characteristic rule by referring to the frequency stored in the DB 210.

As described above, the character-specific sentence generation characteristic rule DB 210 is a database that provides the stored information to the sentence generation processing processor 201 in response to the sentence generation characteristic rule request for the personality information of the sentence generation processing processor 201. , As shown in FIG. 1, it may be a database implemented separately from the sentence generator 200, but in another embodiment, the character is provided together in the sentence generator 200 and according to the request of the sentence generator 200 Sentence generation characteristic rules for each information can be provided.

Alternatively, in another example, the sentence generation characteristic rule DB 210 for each personality may be implemented as a server separate from the sentence generator 200, and the sentence generator 200 is in the sentence generation characteristic rule DB 210 for each personality. By accessing and requesting a sentence generation characteristic rule for each personality information, the sentence generator 200 may receive a sentence generation characteristic rule for each personality information from the server accordingly, but is not limited thereto.

At this time, in an example in which the character-specific sentence generation characteristic rule DB 210 is implemented as a server, the character-specific sentence generation characteristic rule DB 210 transmits and receives signals and data through wired communication or wireless communication with the sentence generator 200 It may include a wired communication module or a wireless communication module.

Continuing to describe the sentence generation processing processor 201, the sentence generation processing processor 201 is transmitted from the word selector 100 based on the sentence generation characteristic rule transmitted from the sentence generation characteristic rule DB 210 for each personality. A sentence is automatically generated according to the selected word and input personality information 10 and transmitted to the dialogue generator 300.

In one example, when the selected word transmitted from the word selector 100 is'yes' and the input personality information 10 is a positive personality, the sentence generation processing processor 201 is ), which is a character-specific sentence generation characteristic rule for positive personality information, is provided with a'friendly and long sentence type', and uses them to generate sentences to be used in the next dialogue following the previous dialogue (20). In this embodiment, the sentence generation processing processor 201 is'Yes. I ate it.’

In this case, when the word selector 100 selects a plurality of words, the sentence generation processing processor 201 may generate a sentence using the plurality of selected words, but is not limited thereto.

The sentence generation processing processor 201 may transmit the generated sentence to the conversation generator 300 through the communication module 202. In this case, the communication module 202 may be a communication module that performs wired or wireless communication with the communication module 102 and the dialog generator 300 of the word selector 100. The communication module 202 may communicate with the character-specific sentence generation characteristic rule DB 210, or may access the character-specific sentence generation characteristic rule DB 210 formed as a server.

In addition, the sentence generation processing processor 201 may store the generated sentence in the storage device 203. At this time, the storage device 203 may be a volatile memory or a nonvolatile memory that stores the sentence generated by the sentence generation processing processor 201, and is generated by receiving the personality information 10 together from the sentence generation processing processor 201 You can match and store a sentence.

In addition, the sentence generation processing processor 201 may output the generated word to an output device of the input/output device 204. The input/output device 204 also includes an input device that receives a signal for controlling to update the stored data of the character-specific sentence generation characteristic rule DB 210.

In one example, the input device may be a mouse or a keyboard, and the output device may be a monitor or speaker, but is not limited thereto.

Again, referring to FIG. 1, when the conversation generator 300 of the conversation generator S using personality information is described, the conversation generator 300 is connected to the sentence combination characteristic rule DB 310 for each personality, and the sentence generator ( A device that receives a sentence generated for personality information from 200) and generates a next conversation to be provided after the previous conversation 20 initially input to the word selector 100, as shown in FIG. 4, a conversation generation processing processor 301, a communication module 302, a storage device 303, and an input/output device 304 may be included.

The conversation generation processing processor 301 of the conversation generator 300 automatically creates a conversation for each personality by referring to the personality information 10, the sentence transmitted from the sentence generator 200, and the sentence combination characteristic rule DB 310 for each personality. Is the generating processor.

The sentence combination characteristic rule DB 310 for each personality referenced by the conversation generation processing processor 301 may be provided by matching sentence combination rules for personality information as shown in Table 3 below.

Personality information Characteristic Sentence Combination Characteristics Rules positive Provided after the interval between previous conversations negative Provided right after previous conversation

The personality information included in the character-specific sentence combination characteristic rule DB 310 may further include optimistic, pessimistic, extroverted, introspective, emotional, depressed personality, and capricious personality, in addition to positive and negative personality, and may further include sentence combination characteristic rule DB for each personality. The number and types of personality information included in (310) are not limited.

The character-specific sentence combination characteristic rule DB 310 may include a plurality of character-specific sentence combination characteristic rules for one personality information, and may be updated to be added or modified according to an input signal. When the character-specific sentence combination characteristic rule DB 310 includes a plurality of character-specific sentence combination characteristic rules for one personality information, the frequency referenced by the conversation generation processing processor 301 may be matched to each rule and stored. , The conversation generation processing processor 301 may extract a sentence combination characteristic rule with reference to the frequency stored in the DB 310.

In one example, the character-specific sentence combination characteristic rule DB 310 is set to have a conversation rule by connecting sentences according to a causal relationship in a specific personality, and randomly connecting sentences in order not to match the causal relationship in another specific personality. You may be setting it to have rules for conversation.

As described above, the character-specific sentence combination characteristic rule DB 310 is a database that provides the stored information to the conversation generation processing processor 301 in response to a sentence combination characteristic rule request for the personality information of the conversation generation processing processor 301. , As shown in FIG. 1, it may be a database implemented separately from the sentence generator 200, but in another embodiment, it is provided with the inside of the conversation generator 300 to respond to the request of the conversation generating processor 301 Accordingly, a sentence combination characteristic rule for each personality information can be provided.

Alternatively, in another example, the character-specific sentence combination characteristic rule DB 310 may be implemented as a separate server from the dialogue generator 300, and the dialogue generator 300 is used in the character-specific sentence combination characteristic rule DB 310. By accessing and requesting a sentence combination characteristic rule for each personality information, the conversation generator 300 may receive a sentence combination characteristic rule for each personality information from the server accordingly, but is not limited thereto.

At this time, in an example in which the character-specific sentence combination characteristic rule DB 310 is implemented as a server, the character-specific sentence combination characteristic rule DB 310 transmits and receives signals and data through wired communication or wireless communication with the conversation generator 300 It may include a wired communication module or a wireless communication module.

Continuing the description of the conversation generation processing processor 301, the conversation generation processing processor 301 is based on the sentence combination characteristic rule received from the character-specific sentence combination characteristic rule DB 310, and the input personality information 10 Accordingly, sentences received from the sentence generator 200 are combined.

In one example, when the input personality information 10 has a positive personality, the conversation generation processing processor 301 refers to the sentence combination characteristic rule DB 310 for each personality, and from the sentence generator 200 after the previous inter-conversation interval. At least one sentence received is provided, and a dialogue is created by combining a plurality of sentences.

The conversation generation processing processor 301 may transmit the conversation generated according to the sentence combination rule to an external server through the communication module 302. In this case, the external server may be a game server mentioned in the description of the word selector 100. The communication module 302 may be a communication module that performs wired or wireless communication with the communication module 202 of the sentence generator 200.

In addition, the conversation generation processing processor 301 may store the generated conversation in the storage device 303. At this time, the storage device 303 may be a volatile memory or a nonvolatile memory that stores the conversation generated by the conversation generation processing processor 301, and is generated by receiving the personality information 10 together from the conversation generation processing processor 301 Match one conversation and save it.

In addition, the conversation generation processing processor 301 may output the generated conversation to an output device of the input/output device 304. The input/output device 304 also includes an input device that receives a signal for controlling to update the stored data of the sentence combination characteristic rule DB 310 for each personality.

In one example, the input device may be a mouse or a keyboard, and the output device may be a monitor or speaker, but is not limited thereto.

In the conversation generator S using personality information, the communication modules 102, 202, 302 of the word selector 100, the sentence generator 200, and the conversation generator 300 are transmitted through an external server, or a word Personality information list stored in DBs 110, 210, 310 in which personality information input through input/output devices 104, 204, and 304 of the selector 100, sentence generator 200, and conversation generator 300 If not present, the processors 101, 201, 301 replace the corresponding personality information with one of the most similar personality information among the personality information defined in the DBs 110, 210, 310, and the replaced personality information Use to select a word, or create a sentence and a dialogue.

In an example in which the processors 101, 201, 301 are not present in the DBs 110, 210, 310 of the received personality information, and replace them with personality information existing in the DBs 110, 210, 310, a word The storage devices 103, 203, and 303 of the selector 100, the sentence generator 200, and the conversation generator 300 store similar personality information with respect to the personality information defined in the DBs 110, 210, 310. Match and save.

For example, the storage devices 103, 203, and 303, as shown in FIG. 5, are similar personality information about the positive personality defined in the DBs 110, 210, 310, optimistic personality, optimistic personality, and favorable personality. You may be storing your personality by matching it.

Accordingly, the'optimistic personality' which is the personality information transmitted from the communication modules 102, 202, 303 or the input/output devices 104, 204, 304 by the processors 101, 201, 301 is converted into the DBs 110, 210, 310), the'positive personality' matching the'optimistic personality' with similar personality information from the matched similar personality information as shown in FIG. 5 is extracted as alternative personality information, and the extracted'positive personality' Is searched in the DBs 110, 210, 310, and word selection, sentence generation, and dialogue generation processing are performed using rule information matched thereto.

In another example, the processors 101, 201, and 301 are DBs 110, 210, and the personality information transmitted through the communication modules 102, 202, and 302 or input/output devices 104, 204, and 304 If not defined in 310), similar personality information and similar personality information are searched by accessing the Internet through the communication modules 102, 202, 302 and searching for the corresponding personality information not defined in the DBs 110, 210, 310. Among them, one of the personality information defined in the DBs 110, 210, 310 can be extracted and used as alternative personality information to be used for word selection, sentence generation, and dialogue generation processing.

The conversation generator using personality information according to another embodiment of the present invention with reference to FIG. 6 includes a word selector 100 for selecting a word by referring to a character-specific word use characteristic rule DB 110, and a word selector 100. ) And a conversation with reference to the character-specific word combination characteristic rule DB 320 may be implemented by including the conversation generator 300a.

In this case, the word selector 100 and the character-specific word use characteristic rule DB 110 are the same as those described with reference to FIGS. 1 to 2 and 5, and are not described in detail in this embodiment, but in the same context. It should be understood.

However, in this case, the dialogue generator 300a in this embodiment with reference to FIG. 6 receives the word selected by the word selector 100 from the word selector 100, and refers to the word combination characteristic rule DB 320 for each personality. Conversation is created by combining the words received.

The conversation generator 300a may be implemented in the structure shown in FIG. 4, and the conversation generation processing processor 301 includes the personality information 10, the word transmitted from the word selector 100, and the combination characteristic rule DB for each personality ( 320), it is a processor that automatically generates a dialogue by personality.

The character-specific word combination characteristic rule DB 320 referenced by the conversation generation processing processor 301 may be provided by matching character-specific word combination characteristic rules for personality information, like the integrated DBs of Tables 2 and 3.

The personality information included in the character-specific word combination characteristic rule DB 320 may include a plurality of personality information as described with reference to FIGS. 1 to 5, and is not limited thereto, and may be added or modified according to an input signal. Can be updated. In addition, the character-specific word combination characteristic rule DB 320 matches and stores the frequency referenced by the conversation generation processing processor 301, and the conversation generation processing processor 301 creates a conversation with reference to the frequency stored in the DB 320. It is possible to extract the word combination characteristic rule for

Characteristic word combination characteristic rule DB 320 is a database that provides information stored at the request of the conversation generation processing processor 301 to the conversation generation processing processor 301, as shown in Fig. 6, the conversation generator ( It may be a database implemented separately from 300a), but in another embodiment, it is provided with the inside of the conversation generator 300a to provide a word combination characteristic rule for each personality information at the request of the conversation generation processing processor 301. I can.

Alternatively, in another example, the character-specific word combination characteristic rule DB 320 may be implemented as a separate server from the conversation generator 300a, and the conversation generator 300a is provided in the character-specific word combination characteristic rule DB 320. By accessing and requesting information, the conversation generator 300a may receive a word combination characteristic rule for each personality information from the server, but is not limited thereto.

At this time, in an example in which the character-specific word combination characteristic rule DB 320 is implemented as a server, the character-specific word combination characteristic rule DB 320 transmits and receives signals and data through wired communication or wireless communication with the conversation generator 300a. It may include a wired communication module or a wireless communication module.

In addition, the communication module 302 of the conversation generator 300a may be a communication module that performs wired or wireless communication with the communication module 102 of the word generator 100. In the embodiment with reference to FIG. 6, the communication module 102 of the word generator 100 may be a communication module that performs wired or wireless communication with the communication module 302 of the conversation generator 300a.

The input/output device 304 may include an input device that receives a signal for controlling to update the stored data of the word combination characteristic rule DB 320 for each personality.

Unlike in the embodiment referring to FIG. 1 that the conversation generator 300 generates a conversation using a sentence generated by the sentence generator 200, the conversation generator 300a in an example referring to FIG. It is formed in a structure in which a conversation is generated using a word selected by the word selector 100, and a conversation can be automatically generated according to personality information.

In addition, in the embodiment shown in FIG. 7, the conversation generator S using personality information includes both a word selector 100, a sentence generator 200, and a conversation generator 300a, and the word selector 100 A word is selected, and the sentence generator 200 is connected to the word selector 100 to generate a sentence using the word selected by the word selector 100. The dialogue generator 300a is connected to the word selector 100 and the sentence generator 200, respectively, but basically, a dialogue is generated using the word selected by the word selector 100, and the sentence generated by the sentence generator 200 is It may be transmitted from the sentence generator 200 and referred to for generating a conversation. In this case, the conversation generator 300a may request the sentence generator 200 to generate a sentence to receive the sentence or may request the sentence generated by the sentence generator 200 to receive the sentence, but is not limited thereto.

In this way, the conversation generator S using personality information automatically generates a conversation to be provided after the previous conversation 20 according to the input personality information 10, so a conventional conversation generator that generates only simple conversational sentences such as a counselor. There is an effect that can overcome the limitations of

The conversation generator S using personality information according to the present embodiment described with reference to FIGS. 1 to 7 is applied to an app or software for personal assistant, a chat program, or generates dialogue of characters such as movies, dramas, and animations. Can be used to do. Alternatively, it can be used for game NPCs, game monsters, or kiosks, and the scope of application should not be limited.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements by those skilled in the art using the basic concept of the present invention defined in the following claims are also provided. It belongs to the scope of rights.

S: Conversation generator using personality information 10: Personality information
20: previous conversation 100: word picker
110: Characteristic rules for using words by personality 200: Sentence generator
210: Characteristic rules for generating sentences by personality 300: Conversation generator
310: Characteristic Sentence Combination Characteristics Rule DB

Claims (10)

Characteristic word use characteristic rule DB that matches and stores at least one personality word with respect to at least one personality information,
A word selector that is connected to the personality-specific word use characteristic rule DB and selects a personality-specific word by using the personality information-specific words stored in the personality-specific word use characteristic rule DB according to the input previous conversation content and input personality information ,
Characteristic sentence generation characteristic rule DB that matches and stores at least one character-specific sentence generation characteristic rule with respect to at least one personality information, and,
A sentence generator that is connected to the character-specific sentence generation characteristic rule DB and generates a sentence by combining a word selected from the word selector according to input personality information,
A sentence combination characteristic rule DB for each personality that matches and stores at least one sentence combination characteristic rule for each of the positive and negative personality information among at least one personality information, and connects to the sentence combination characteristic rule DB for each personality And, according to the input personality information, further comprising a dialogue generator for generating a dialogue by combining sentences using the sentence combination characteristic rule for each personality information stored in the sentence combination characteristic rule DB for each personality,
The word usage characteristic rule DB for each personality is
Personality information is classified into positive personality and negative personality based on the words of the stored personality information and previous conversation contents, and the word selection processing processor is provided to match and store the referenced frequency and exclamation with the word information for each personality,
The sentence generation characteristic rule DB for each personality is
If the personality information is positive, it is set in a concise and long sentence type, and if it is a negative personality, it is provided in a short answer type
The sentence combination characteristic rule DB for each personality is
Personality, characterized in that provided to provide a conversation to be provided after the previous conversation content after an interval between previous conversations based on the input personality information when the personality information is positive, and to provide immediately after the previous conversation when the personality information is negative Conversation generator using information.
The method of claim 1,
The word selector,
A word selection processing processor that selects a word to be used for a next conversation according to the previous conversation, but searches for the input personality information in the character-specific word use characteristic rule DB to select a word for the input personality information, and
Communication module for transmitting the word selected by the word selection processing processor to the outside, or receiving the previous conversation content or the personality information and transmitting it to the word selection processing processor
Including,
The sentence generator,
A sentence generation processing processor for generating a sentence using a word selected by the word selection processing processor by searching for the input personality information in the sentence generation characteristic rule DB for each personality, and
Sending the sentence generated by the sentence generation processing processor to the outside, receiving the selected word from the communication module of the word selector and transmitting it to the sentence generation processing processor, or transmitting the sentence generated by the sentence generation processing processor to the outside Communication module
Conversation generator using personality information, characterized in that it comprises a.
The method of claim 2,
The word selector,
A storage device for storing the word selected by the word selection processing processor, and
An input/output device including an output device for outputting a word selected by the word selection processor, and an input device for receiving the previous conversation content or the personality information and transmitting the input to the word selection processor
Including more,
The sentence generator,
A storage device that stores a sentence generated by the sentence generation processing processor, and
An input/output device including an output device for outputting a sentence generated by the sentence generation processing processor and an input device for receiving the personality information and transmitting the received personality information to the sentence generation processing processor
Conversation generator using personality information, characterized in that it further comprises.
The method of claim 1,
The personality-specific word use characteristic rule DB is updated by receiving the personality information and words for each personality information additionally.
delete delete The method of claim 1,
The conversation generator,
A conversation generation processing processor for generating a conversation for the input personality information by searching for the input personality information in the sentence combination characteristic rule DB for each personality, and
A communication module that transmits the conversation generated by the conversation generation processing processor to the outside or receives the personality information to the conversation generation processing processor
Conversation generator using personality information, characterized in that it comprises a.
The method of claim 7,
The conversation generator,
A storage device for storing the conversation generated by the conversation generation processing processor, and
An input/output device including an output device for outputting a dialog generated by the dialog generation processing processor and an input device for receiving the personality information and transmitting the received personality information to the dialog generation processing processor
Conversation generator using personality information, characterized in that it further comprises.
The method of claim 3,
The storage device of the word selector and the storage device of the sentence generator,
A conversation generator using personality information, characterized in that matching and storing personality information stored in the character-specific word use characteristic rule DB and the personality information stored in the personality-specific sentence generation characteristic rule DB.
The method of claim 9,
The word selection processing processor and the sentence generation processing processor,
If the input personality information does not exist in the character-specific word use characteristic rule DB or the personality-specific sentence generation characteristic rule DB, search for similar personality information stored in the storage device to extract alternative personality information, and the extracted A conversation generator using personality information, characterized in that a word is selected or a sentence is generated by searching for replacement personality information in the character-specific word use characteristic rule DB or the personality-specific sentence generation characteristic rule DB.

Images