JP6551852B1 - Information providing apparatus, information providing method, and program - Google Patents

Abstract

An object of the present invention is to improve the continuity of a dialog in a system that provides information in a dialog format. An information providing apparatus provides information in a conversational form between a virtual personality and a user by allowing the user to select his / her own dialogue. The information providing device reads a scenario from a storage device that stores a conversational scenario including a virtual personality dialogue and a user dialogue, outputs a virtual personality dialogue and a user dialogue option, and from the user Between the speech control unit that accepts the selection of dialogue, the virtual personality spoken according to the scenario, and the dialogue of the user's dialogue, between the dialogue of the selected user and the speech of the virtual personality An insertion control unit that inserts a filler or a word that expresses gratitude, and the insertion control unit has a probability of inserting a filler or a word that expresses gratitude, and the word that expresses the filler or gratitude It is adjusted so that it rises as the number of utterances from the previous insertion increases. [Selection] Figure 2

Description

  The present invention relates to an information providing apparatus, an information providing method, and a program.

  Conventionally, various chatbots and software agents have been proposed which perform information provision, conversations and the like with users.

  Also, for the purpose of realizing natural speech in speech synthesis, a technology of inserting a redundant expression language or the like representing a redundant expression in speech into the text of a spoken word included in the text data of the speech synthesis target and outputting the speech Have been proposed (e.g., Patent Document 1). Specifically, it is described that a redundant expression language is inserted based on an insertion rule using probability. In addition, as redundant expression languages, fillers such as “e”, ending redundant words such as “~ is”, interjections such as “ma”, sloppy “(silence)”, synonym dictionary data and synonym dictionary data are used. There is a statement that same words and synonyms are registered in the dictionary.

JP 2014-48443 A

  Conventionally, chatbots and software agents that interact with text can not express nuances depending on how they are spoken, unlike spoken language, and there has been no example that actively incorporates expressions such as fillers.

  On the other hand, the inventor of the present application has found that redundant expressions such as fillers are not necessarily useless in order to make users use chat bots and software agents. An object of the present invention is to improve the continuity of interaction in a system for providing information to a user interactively.

  The information providing apparatus provides information in a form of conversation between the virtual personality and the user by allowing the user to select his / her own dialogue. In addition, the information providing apparatus reads a scenario from a storage device storing a dialog type scenario including a dialogue of a virtual personality and a dialogue of the user, and outputs a dialogue of the virtual personality and a choice of the dialogue of the user. A speech control unit that accepts selection of a line from the user, and a speech between the user's speech and the virtual personality, the speech of the selected user based on the history of the virtual personality and the speech of the user who spoke according to the scenario And the insertion control unit for inserting a filler representing a feeling of gratitude or a filler representing a gap between words, and the insertion control unit has a probability of inserting the filler or a feeling of gratitude representing the filler or the feeling of gratitude Adjust so that the number of utterances from the previous insertion of the word to represent increases.

  In this way, fillers and words of thanks can be inserted at preferred positions in a predetermined scenario, and the continuity of the dialogue can be improved. In addition, it is possible to change the same scenario by probabilistically determining the insertion of fillers and words of thanks.

  Further, the insertion control unit analyzes the type of emotion related to the sentence of the speech using a predetermined feature vector determined in association with the word included in the history of the speech, and associates the type of emotion in advance with the type of the emotion. It is also possible to insert a filler or words of appreciation that are stored in the storage device. In this way, it is possible to select a filler or a word expressing gratitude in the context.

  Also, based on the history of the dialogue, the relationship between the presence or absence of a filler or a word expressing gratitude, and the number of series of utterances in the history are statistically analyzed to increase the number of series of utterances in the conversation, The information processing apparatus may further include a learning processing unit that adjusts the probability of inserting a filler or a word expressing gratitude. By preferentially selecting words and timings that increase the number of dialogues, it is possible to realize the insertion of words in a manner that makes the sense of incongruity less, and to further increase the number of dialogues.

  The contents described in the means for solving the problems can be combined as much as possible without departing from the problems and technical ideas of the present invention. The contents of the means for solving the problems can be provided as a device such as a computer or a system including a plurality of devices, a method executed by the computer, or a program executed by the computer. Note that a recording medium for holding the program may be provided.

  In a system that provides information to the user interactively, the continuity of the dialogue can be improved.

It is a figure which shows an example of a structure of the whole system. It is a block diagram which shows an example of a structure of a user apparatus and an interactive information provision apparatus. It is a figure which shows an example of the image displayed on a user apparatus. It is the figure which represented visually an example of the scenario memorize | stored in the interactive information provision apparatus. It is a figure which shows an example of the output at the time of inserting a filler in the interface of a scenario. It is a processing flow figure showing an example of conversation offer processing. It is a processing flow figure showing an example of conversation offer processing. It is a figure which shows an example of the filler or the word of thanks memorize | stored in a memory | storage device. It is a processing flow figure showing an example of conversation offer processing concerning a modification. It is a figure which shows an example of the image displayed in a modification. It is a processing flow figure showing an example of conversation offer processing concerning a modification. It is a figure which shows an example of the image displayed in a modification.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

<System configuration>
FIG. 1 is a diagram illustrating an example of the entire system according to the embodiment. This system provides a software agent, which is a virtual personality in which a speech is output by a computer and has a conversation with a user. The content of the conversation proceeds based on information representing a scenario created in advance.

This system includes a user device 1 and an interactive information providing device 2, which are connected via a network 3. The network 3 is a communication network such as the Internet, and computers connected to the network 3 can communicate with each other based on various protocols. The user device 1 is a smartphone, tablet, PC (Personal Computer
And the like, and functions as a user interface of the conversation providing system according to the present embodiment by executing a program (also referred to as software or application) according to the present embodiment. Note that a plurality of user devices 1 may be connected to the network 3. The interactive information providing device 2 is a general computer, and outputs a statement of a virtual personality to the user device 1 through the network 3 by executing the program according to the present embodiment. The process of obtaining the user response from is repeated. The user's speech is input to the user device 1 as the user selects from the options output from the interactive information providing device 2, and is transmitted to the interactive information providing device 2 via the network 3.

<Device configuration>
FIG. 2 is a block diagram illustrating an example of the configuration of the user device 1 and the interactive information providing device 2.

The user device 1 is a general computer such as a smartphone or a tablet, and has an input / output interface (I / F) 11, a storage device 12, a communication interface (I / F) 13, a processor 14, and a bus 15. It has. The input / output I / F 11 is a user interface such as a touch panel. The input / output I / F 11 displays an utterance of a virtual personality to the user, displays an option of the user's response, and receives a selection of the user's response based on the user's operation. The storage device 12 is a main storage device such as a random access memory (RAM) or a read only memory (ROM), and an auxiliary storage device (a secondary storage device such as a hard disk drive (HDD), a solid state drive (SSD)) or a flash memory. Device). The main storage device temporarily stores programs read by the processor and data of received speech, and secures a work area of the processor. The auxiliary storage device stores a program executed by the processor. The communication I / F 13 is a network module that performs communication with, for example, a mobile phone base station or a wireless LAN (Local Area Network) access point, and performs communication based on a predetermined protocol. The processor 14 is an arithmetic processing unit such as a CPU (Central Processing Unit), and performs each process according to the present embodiment by executing a program. In the example of FIG. 2, functional blocks are shown in the processor 14. Specifically, the processor 14 functions as the input / output control unit 141. The input / output control unit 141 displays an utterance of a virtual personality on the input / output I / F 11 and accepts a user response input via the input / output I / F 11. The components as described above are connected via the bus 15.

The interactive information providing apparatus 2 is a general computer such as a server, and includes an input / output I / F 21, a storage device 22, a communication I / F 23, a processor 24, and a bus 25. The input / output I / F 21 is a user interface such as a keyboard, a mouse, or a display. The storage device 22 is a main storage device such as a RAM or a ROM, and an auxiliary storage device such as an HDD, an SSD, or a flash memory. The main storage device temporarily stores programs read by the processor and scenario data stored in the auxiliary storage device, and secures a work area for the processor. The auxiliary storage device stores a program executed by the processor. The communication I / F 23 is a wired network card, for example, and performs communication based on a predetermined protocol. The processor 24 is an arithmetic processing unit such as a CPU, and performs each process according to the present embodiment by executing a program. As illustrated in FIG. 2, the processor 24 functions as a conversation control unit 241, an insertion control unit 242, and a learning processing unit 243. The conversation control unit 241 outputs an utterance of a virtual personality for the user and an option of the user's response to the utterance using a scenario prepared in advance. The insertion control unit 242 determines whether to insert a filler or a word expressing gratitude in accordance with a predetermined rule. The learning processing unit determines whether the number of dialogues with the user has increased by inserting a filler or a word expressing gratitude based on statistical processing, and the insertion control unit 242 generates a word expressing a filler or gratitude Adjust the probability of insertion. The components as described above are connected via the bus 15.

<User interface>
FIG. 3 is a diagram illustrating an example of an image displayed on the input / output I / F 11 of the user device 1. In the example of FIG. 3, an image 111 representing an agent, an image 112 representing a user, a balloon 113 representing an agent's utterance, and a user's response option 114 are displayed on the input / output I / F 11 which is a touch panel. Yes. The user can interact with the agent, for example, by tapping his / her speech from such options. The number of options is one or more, and is not limited to four.

<Scenario>
FIG. 4 is a diagram visually illustrating an example of a scenario stored in advance in the storage device 22 of the interactive information providing apparatus 2. In the present embodiment, processing by software such as agent dialogues or telephone calls represented by rectangles and options of user dialogues represented by rounded rectangles are arranged in order by arrows. The agent dialogues are basically connected in series, and are output to the outlet 113 shown in FIG. The user's choices are displayed in the choice 114 shown in FIG. At this time, the options arranged in parallel are output in parallel, the scenario branches according to the line selected by the user, and the subsequent lines are displayed. In addition, it may be possible to perform the telephone call processing shown in FIG. 4, display of a website (not shown), and the like by using a function provided in the user apparatus.

<Inserting fillers and words of thanks>
FIG. 5 is a diagram illustrating an example of an output when a filler is inserted between a scenario dialogue. The insertion control unit 242 of the interactive information providing apparatus 2 according to the present embodiment has, among the prepared scenarios, “Eh”, “Eh”, “Uneun”, “Hou”, “I see”, etc. Insert words that express gratitude, such as fillers that fill the conversation and “thank you”. Specifically, the agent may insert these words based on a predetermined probability after the opponent's dialogue.

  The inserted words may be stored in advance in association with emotion types such as emotions. Then, for the words included in the other party's speech, the emotion in the flow of the dialogue is analyzed using the feature vector having the emotion type as an element, and the word to be inserted is selected according to the emotion type. In addition, the above-described probability may be increased as the number of conversations (the number of dialogues made by the user and the agent) after the previous filler or the like is inserted increases.

  According to the scenario shown in FIG. 4, after the user's choice "What can you do?", The agent's statement "order item ..." is connected. Is inserted with the filler "Huh ..." based on the conditions described above.

In addition, the conversation control unit 241 of the interactive information providing device 2 stores the log of the interaction between the user and the agent in the storage device 22, and the learning processing unit 243 inserts the filler and the words of thanks by the number of interactions. Machine learning whether or not it has been extended. Then, in the learning processing unit 243, these words are inserted in the words judged that the number of dialogues has been increased by inserting, and in the timings when it is judged that the number of dialogues has increased by inserting (during an interview in a scenario). Feedback is given to increase the probability of insertion. Whether or not the number of conversations has increased is determined based on whether or not the user has advanced the scenario prepared in advance, more than the number of inserted dialogues.

<Conversation processing>
FIG. 6 is a process flow diagram showing an example of the conversation providing process executed in the present system. First, the conversation control unit 241 of the interactive information providing apparatus 2 reads out a scenario as shown in FIG. 4 (FIG. 6: S1). In this step, based on the order defined in the scenario, an agent dialogue or a user dialogue option, or an instruction to perform other processing is read. In the example of FIG. 4, the agent's dialogue “Is shopping?” Connected by an arrow is read from “Start”.

  Then, the conversation control unit 241 determines whether the read-out line is the agent's line (FIG. 6: S2). If it is an agent line (S2: YES), the conversation control unit 241 outputs the read agent line via the input / output I / F 11 (FIG. 6: S3). For example, as shown in FIG. 3, the agent's statement "Do you shop?" Is output. Then, the process transitions to the process of FIG. 7 through the connector “A”.

On the other hand, if the read-out text is not the agent's text (S2: NO), the conversation control unit 241 determines whether the read-out text is an option of the user's text (FIG. 6: S4). If it is not a user's choice of dialogue (S4: NO), the conversation control unit 241 performs processing according to the scenario (FIG. 6: S5). In this step, processing such as calling a predetermined telephone number and displaying a website of a predetermined URI (Uniform Resource Identifier) is executed. And it changes to the process of FIG. 7 via connector "A".

  On the other hand, if the read-out dialogue is an option of the user's dialogue (S2: YES), the conversation control unit 241 displays the option of the user's dialogue (FIG. 6: S6). For example, in the scenario shown in FIG. 4, at the next stage of the agent's statement "Shopping?", The user's option "I want to see the product", "What can I do?" "I want to call" is read out, and is output as shown in option 118 of FIG. 3 in this step.

  Thereafter, the conversation control unit 241 receives the selection of the user via the input / output I / F 21 (FIG. 6: S7). In this step, for example, one of the choices of dialogue is selected according to the position where the user taps the touch panel. Then, the process transitions to the process of FIG. 7 via the connector “B”.

  FIG. 7 is a process flow diagram showing an example of a conversation providing process performed in the present system. After S7, the insertion control unit 242 of the interactive information providing apparatus 2 determines whether to insert a filler or a word of thanks (FIG. 7: S8). In this step, it is determined to insert a filler or a word of thanks based on a predetermined probability. The history of the speech of the agent or user who spoke according to the scenario may be stored in the storage device 22, and the probability may increase as the number of interactions increases from the previous insertion.

  Also, words to be inserted may be selected based on the history. For example, by morphological analysis, it is decomposed into words included in the sentence that has been spoken and analyzed using feature vectors whose elements are emotion types. For example, the insertion control unit 242 may use Word2Vec, Doc2Vec, or the like to perform emotion analysis based on words included in the history of the dialogue of the agent or user. Then, the ratio of the emotion types “joy”, “anger”, “sorrow”, and “easy” is obtained. Also, based on the proportions of emotions, the probability of selecting each of the words to be inserted is further adjusted.

FIG. 8 is a diagram illustrating an example of fillers or words of thanks stored in the storage device 22. For example, words stored in the storage device 22 in advance associated with the emotion type are inserted. The table of FIG. 8 has the attributes of emotions and attributes of filler or thanks words. In the emotion type field, “joy”, “anger”, “sorrow”, and “easy” are registered. Also, in the field of the filler or the word of thanks, words such as "uh", "un-un", "love", "thank you" are registered. As a result of the emotion analysis described above, the probability that a filler or a word of thanks corresponding to the emotion corresponding to the conversation history is selected is increased. A plurality of tables as shown in FIG. 8 may be prepared depending on the nature of the agent.

  In addition, the learning processing unit 243 determines whether the number of dialogues has increased by inserting based on the history of speech of the agent or the user, or at the timing when it is determined that the number of dialogues has increased by inserting. Alternatively, it may be further adjusted to increase the probability that a word of thanks is inserted. The learning process increases the probability that the word is selected, for example, as the difference between the expected number of dialogues is larger, based on the conditional probability in the case where a certain word is inserted and the case where it is not obtained. It may be. Also, based on the conditional probability in the case where the filler or the word of thanks is inserted at a certain timing and the case where it is not so, the larger the difference between the expected number of dialogues, the higher the probability that the word is inserted at that timing. It may be. Further, regarding the combination of the word selection and the timing selection, the probability that the combination is selected may be increased as the number of conversations expected from the history increases.

  If it is determined that a filler or a word of thanks is to be inserted (S8: YES), the insertion control unit 242 outputs the selected dialogue (FIG. 7: S9). In this step, for example, as shown in FIG. 5, fillers or words of appreciation that were not in the scenario are output.

  On the other hand, when it is determined that the filler or the word of thanks is not inserted (S8: NO), or after S3 in FIG. 6, the conversation control unit 241 determines whether the scenario read in S1 is ended (FIG. 7: S10). In this step, for example, in the scenario as shown in FIG. 5, it is determined that the scenario has been completed when the end of a line or the like connected by an arrow is reached.

  When the scenario ends (S10: YES), the conversation control unit 241 ends the conversation providing process. On the other hand, if the scenario has not ended (S10: NO), or after S9, the process returns to S1 in FIG. 6 via the connector “C”.

<Effect>
Especially in chat-type software agents, fillers tend to be considered redundant and useless. However, the inventors found that after inserting a filler or a word of gratitude, the number of interactions with the user is increased even in the same scenario through an experiment on chat-type agents. An increase in the number of dialogues means an increase in usage time (active rate) of the software, which is considered to increase user satisfaction and happiness.

  According to the above-described embodiment, fillers and words of gratitude can be inserted at preferred positions in a predetermined scenario, and the continuity of the dialogue can be improved. In addition, it is possible to change the same scenario by probabilistically determining the insertion of fillers and words of thanks. Also, by performing learning processing on the history of conversation and selecting words and timings that increase the number of conversations, it is possible to realize the insertion of words with less discomfort, and further extend the number of conversations.

<Modification>
When there are a plurality of agents, a filler or words expressing gratitude can be inserted in the same manner. In this modification, the role is divided into an agent that speaks a dialog according to a scenario and an agent who speaks a filler or a word of thanks.

  FIG. 9 is a process flow diagram showing an example of a conversation providing process according to a modification. This modification is basically the same as the process of FIG. 6, and corresponding steps are denoted by the same reference numerals. Hereinafter, differences from FIG. 6 will be mainly described.

  In the modified example of FIG. 9, it is determined whether or not it is a line of one agent (agent 1) in S2, and the line of agent 1 is output in S3.

  FIG. 10 is a view showing an example of an image displayed in the modification. In the image of FIG. 10, in addition to the first agent 111 (agent 1) also shown in FIG. 3, the second agent 115 (agent 2) is displayed.

  Further, after S3, the process transits to S8 of FIG. 11 via the connector "B". FIG. 11 is a process flow diagram showing an example of a conversation providing process according to a modification. That is, in S8, it is determined whether the other agent (agent 2) speaks the filler or thank-you not only after the user speaks but also after the agent 1 speaks. If it is determined to insert a filler or a word of thanks (S8: YES), the agent 2 outputs the selected dialogue in S9.

  FIG. 12 is a view showing an example of an image displayed in the modification. In the example of FIG. 12, a balloon 116 representing the message of the agent 2 is displayed. Also, in FIG. 12, after the agent 1's remarks, the agent 2's dialogue “How” is inserted. Thus, the agent 2 speaks a filler or a word of thanks after the agent 1 or after the user. The probability of inserting a word can be determined in the same manner as in the above-described embodiment.

  In this way, the number of dialogues with the user can be extended even in three-way conversations.

  Also, in the above-described variation, although the roles are divided into an agent who speaks a dialogue according to a scenario and an agent who speaks a filler or a word of thanks, the present invention is not limited to such an example. A scenario in which dialogues are assigned to two agents in advance may be used. In this case, for example, one agent may say a word of filler or gratitude to the other agent and the user. Also, the other agent speaks a filler or thankful to one agent and the user. At this time, the probability of inserting a word can be determined as in the above embodiment.

<Others>
The configuration of the above-described embodiment is an exemplification, and can be combined or changed as much as possible without departing from the subject and technical idea of the present invention.

  The present invention includes a computer program that performs the above-described processing. Furthermore, a computer-readable recording medium on which the program is recorded also belongs to the category of the present invention. With respect to the recording medium on which the program is recorded, the above-described processing can be performed by causing the computer to read and execute the program on the recording medium.

Here, the computer-readable recording medium refers to a recording medium in which information such as data and programs is accumulated by electrical, magnetic, optical, mechanical, or chemical action and can be read from a computer. Examples of such a recording medium that can be removed from the computer include a flexible disk, a magneto-optical disk, an optical disk, a magnetic tape, and a memory card. In addition, examples of the recording medium fixed to the computer include a hard disk drive and a ROM.

1: User device 11: Input / output I / F
12: Storage device 13: Communication I / F
14: Processor 141: Input / output control unit 2: Interactive information providing device 21: Input / output I / F
22: Storage device 23: Communication I / F
24: Processor 241: Conversation control unit 242: Insertion control unit 243: Learning processing unit 3: Network

Claims (4)

An information providing apparatus that provides information in a conversational form with a virtual personality and the user by allowing the user to select his / her dialogue,
The scenario is read out from the storage device storing the dialogue type scenario including the dialogue of the virtual personality and the dialogue of the user, and the dialogue of the virtual personality and the option of the dialogue of the user are output, and A conversation control unit that accepts dialogue selection;
It is a word which fills a gap between remarks between the words of the selected user and the lines of the virtual personality based on the virtual personality uttered according to the scenario and the history of the user's words An insertion control unit for inserting a filler or words expressing gratitude,
Based on the history of the line, the relationship between the presence or absence of the filler or the word expressing the feeling of gratitude, and the number of the series of utterances in the history is statistically analyzed to increase the number of utterances in the conversation , A learning processing unit that adjusts the probability of inserting the filler or words expressing gratitude,
Equipped with
The insertion control unit adjusts the probability that the filler or a word expressing gratitude is increased as the number of utterances from when the filler or a word expressing gratitude is inserted increases. Information providing device. The insertion control unit analyzes the type of emotion related to the sentence of the speech using a predetermined feature vector determined in association with the word included in the history of the speech, and corresponds in advance to the type of the emotion The information providing device according to claim 1, wherein a word indicating the feeling of gratitude or the filler that is attached and stored in the storage device is inserted. An information providing method in which a computer provides information in a conversational form with a virtual personality and the user by allowing the user to select his own line,
The scenario is read out from the storage device storing the dialogue type scenario including the dialogue of the virtual personality and the dialogue of the user, and the dialogue of the virtual personality and the option of the dialogue of the user are output, and Accept selection of dialogue,
It is a word which fills a gap between remarks between the words of the selected user and the lines of the virtual personality based on the virtual personality uttered according to the scenario and the history of the user's words Insert filler or gratitude words,
Based on the history of the line, the relationship between the presence or absence of the filler or the word expressing the feeling of gratitude, and the number of the series of utterances in the history is statistically analyzed to increase the number of utterances in the conversation , Adjust the probability of inserting the filler or words expressing gratitude,
As for the insertion of the filler or the word expressing gratitude, the probability of inserting the filler or the word expressing gratitude is increased as the number of utterances from the previous insertion of the filler or gratitude is increased. Information provision method to adjust so that By allowing the user to select his own line, a computer that provides information in a conversational form with the virtual personality and the user,
The scenario is read out from the storage device storing the dialogue type scenario including the dialogue of the virtual personality and the dialogue of the user, and the dialogue of the virtual personality and the option of the dialogue of the user are output, and Accept selection of dialogue,
It is a word which fills a gap between remarks between the words of the selected user and the lines of the virtual personality based on the virtual personality uttered according to the scenario and the history of the user's words Insert filler or gratitude words,
Based on the history of the line, the relationship between the presence or absence of the filler or the word expressing the feeling of gratitude, and the number of the series of utterances in the history is statistically analyzed to increase the number of utterances in the conversation , Adjust the probability of inserting the filler or words expressing gratitude,
As for the insertion of the filler or the word expressing gratitude, the probability of inserting the filler or the word expressing gratitude is increased as the number of utterances from the previous insertion of the filler or gratitude is increased. A program that executes a process to adjust so as to rise.

Images