JP2004212664A - Method and device for interaction and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method and a device for interaction that can flexibly determine a next subject according to the interaction result of a new subject when the interaction is carried out for a given purpose according a plurality of predetermined subjects and one subject is interrupted to move to a new subject. <P>SOLUTION: This interaction method comprises: storing unsolved subjects selected for the purpose of interaction with a user in a storage means in the selection order; using the unsolved subject stored finally in the storage means for the purpose of interaction to obtain first information; solving a first unsolved subject stored in the storage means by using the first information to obtain second information; deleting the first unsolved subject which has been solved from the storage means; and deleting from the storage means unsolved subjects selected for the purpose of interaction after a second unsolved subject to use the second unsolved subject related to the second information stored in the storage means in a second step for the purpose of next interaction. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an interaction method for interacting with a user in a natural language to achieve a predetermined object, an interaction device using the same, and a program.
[0002]
[Prior art]
2. Description of the Related Art In recent years, research on an interface that accepts text input by voice or natural language input has been actively conducted. In addition, many expert systems and the like using such an interface have been developed, and a device that accepts input by voice, text, or the like has become available for the general public. In particular, voice input is considered to be easy for the user to use, and a system that accepts voice input using voice recognition technology has been developed.
[0003]
In the case of natural language input by voice, the user rarely speaks all the conditions to be input to the system at once, and requires voice communication between the user and the system. For example, when there is a defect in the user's input, it is necessary to perform a process such as inquiring about the missing condition and integrating the user's answer to it. For this integration process, it is indispensable for the system to grasp the contents of the dialogue currently being performed.
[0004]
As a conventional dialogue management technology, topics are managed in a stack structure, and stack operations are determined to require topic transitions such as "input advances the current topic" or "interruption of the current topic". There is a method of performing based on a result (for example, see Non-Patent Document 1).
[0005]
In this method, the possibility of topic transition is classified according to the dialogue situation based on the number of database search result candidates, and if a topic transition that cannot occur in that dialogue situation is found, the topic It has been determined that the interruption has occurred. When it is determined that an utterance input of a topic has been interrupted, the topic at that time is marked and a new topic is pushed onto the stack to interrupt the topic. When the interrupted topic ends, it is possible to return to the original conversation by returning to the marked topic.
[0006]
However, in the conventional method, when the topic is determined to be interrupted, a mark is added to the topic of the return destination in advance, and when the interrupted topic ends, the mark returns to the mark location, so the interrupted topic contributes to the original topic There is a problem that it is difficult to control the dialogue in such a case.
[0007]
For example, the following dialogue between the system (S) and the user (U) in the car navigation system will be described as an example.
[0008]
S (1): "Determine the destination. Please specify the location."
U (1): "The destination address is Kanagawa Prefecture."
S (2): "Please take your municipal name"
U (2): "After all, by registered name"
S (3): "Please enter your registered name"
U (3): "Yamashita Park"
Here, in the conventional method, the “address determination dialogue” is marked and the “search dialogue by registered name” is pushed in response to the user's second utterance U (2). For this reason, when the search dialogue is completed with the registered name in the third utterance U (3) of the user, the response is as follows to return to the marked address determination dialogue.
[0009]
S (4): “Yamashita Park is here (with map display). Please enter the name of city, town, and village.” As described above, there is a shift in the dialogue intention that the user has selected Yamashita Park as the destination. As a result, satisfactory dialogue cannot be carried out, which imposes a heavy burden on the user.
[0010]
Conventionally, the ambiguity of natural language interpretation has not been considered. In particular, when a voice input is involved, a plurality of recognition candidates appear in the voice recognition result due to ambiguity of recognition. In view of such ambiguity of speech recognition, it is necessary to select a recognition result to be used for a dialog from a plurality of candidates in consideration of the dialog situation at that time. If this countermeasure is not sufficient, the dialogue is advanced on the basis of erroneous recognition, so that a satisfactory dialogue cannot be performed and a great burden is imposed on the user.
[0011]
[Non-patent document 1]
IPSJ Transactions Vol. 9. “Understanding intentions and managing topics in natural language interface systems”
[0012]
[Problems to be solved by the invention]
As described above, conventionally, in order to achieve a predetermined purpose, when a conversation is performed according to a plurality of predetermined topics, the current topic is interrupted, and when a new topic is started, the interrupted original topic is used. After returning to the predetermined topic as the return destination, the user returned to the predetermined topic as the return destination, so even if the goal was achieved as a result of the dialogue with the new topic, Had to talk on another topic.
[0013]
That is, conventionally, when a conversation is performed according to a plurality of predetermined topics (tasks) for achieving a predetermined purpose, when a certain topic is interrupted and shifted to a new topic, the new topic ends. There is a problem that the next topic at that time cannot be flexibly determined according to the result of the conversation with the new topic.
[0014]
Conventionally, the result of voice recognition has been adopted regardless of the dialogue state at that time, and there has been a problem that a satisfactory dialogue cannot be performed and a great burden is imposed on the user.
[0015]
Therefore, in view of the above problems, the present invention interrupts a certain topic (task) and performs a new topic (task) when the dialogue is performed according to a plurality of predetermined topics (tasks) for achieving a predetermined purpose. It is an object of the present invention to provide a dialogue method and apparatus that can flexibly determine the next topic (task) according to the result of a dialogue with the new topic (task) when the process is shifted to ()).
[0016]
It is another object of the present invention to provide a dialogue method and apparatus capable of performing a highly accurate voice dialogue by selecting an optimum candidate from a plurality of candidates obtained as a result of voice recognition according to a dialogue state at that time. .
[0017]
[Means for Solving the Problems]
(1) The present invention is to perform a dialogue with a user to solve a plurality of problems for achieving a predetermined object. The tasks are stored in the storage means in the order of selection, and the unresolved tasks stored last in the storage means are used for the purpose of dialogue, first information is obtained, and the first information is stored in the storage means by the first information. Resolving a first unresolved issue, obtaining second information, deleting the resolved first unresolved issue from the storage means, and rewriting the second information stored in the storage means. In order to use the related second unresolved task in the second step for the next dialogue purpose, delete the unresolved task selected for the dialogue purpose after the second unresolved task from the storage means. As a result, when an issue is interrupted and shifted to a new issue, (By the problems associated with the issues resolved the task output information and following interactive purposes) in accordance with a result of dialogue with the new challenges, it is possible to flexibly determine the next challenge.
[0018]
(2) The second unresolved issue related to the second information satisfies at least one condition for determining that the second information has solved the second unresolved issue.
[0019]
(3) When the second unresolved problem related to the second information is not stored in the storage unit, the first unresolved problem among the unresolved problems stored in the storage unit is The unresolved task selected for the conversation purpose immediately before the selection for the conversation purpose is used for the next said conversation purpose.
[0020]
(4) The storage means is a stack for storing unsolved tasks selected for the purpose of dialogue with the user in the order of selection.
[0021]
(5) When the first unresolved problem solved by the first information is not stored in the storage unit, a third unresolved problem related to the first information is transmitted to the next dialogue. An unresolved issue selected for interactive purposes after the third unresolved issue is deleted from the storage means for use in the purpose, and the storage means stores a third unresolved issue associated with the first information. When the assignment is not stored, a new assignment related to the first information of the plurality of assignments is stored in the storage unit as the unresolved assignment for use in the next conversational purpose.
[0022]
(6) When a plurality of candidates for the first information estimated from the user's input are given, use of the plurality of candidates for the purpose of the dialogue among the unresolved issues stored in the storage means. The candidate related to the unresolved issue that comes later in the use order given priority is selected as the first information.
[0023]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0024]
(1st Embodiment)
According to the present invention, even when it is determined that the topic (dialogue purpose) is changed, the topic is interrupted without previously determining the return destination topic, and the present invention is suitable even when the topic contributes to the original topic. The present invention provides a dialogue system having a function of executing a candidate selection in consideration of a dialogue situation at the time of input by a user. explain.
[0025]
FIG. 1 is a schematic configuration diagram of a dialogue system including a dialogue device according to an embodiment of the present invention, and shows a configuration example of a dialogue system used for a car navigation system, for example. As shown in FIG. 1, the interactive system includes a natural language interpreting apparatus 1, an interactive apparatus 100, and a response output apparatus 2.
[0026]
The interaction device 100 includes a candidate selection unit 101, an interaction stack operation unit 102, an interaction information storage unit 103, and a plan storage unit 104.
[0027]
The natural language interpreting apparatus 1, for example, analyzes a natural language input from a user and outputs an analysis result to the interactive apparatus 100. The input to the natural language interpreting apparatus 1 may be in any form such as text or voice, and an existing interpreting method according to the type can be used. When ambiguity is uttered in the analysis, the candidate and the likelihood are output as a set. Here, a case in which a user's voice input (utterance) is recognized by voice will be described. Usually, a plurality of recognition candidates are obtained by voice recognition. Here, it is assumed that each recognition candidate is paired with the similarity of the input voice as the likelihood and output to the interactive apparatus 100.
[0028]
The interactive device 100 is for interacting with a user in accordance with a plurality of predetermined topics for achieving a predetermined purpose. For example, here, the case where the predetermined purpose is “destination setting” in a car navigation system will be described as an example. That is, the interactive device 100 stores, in the plan storage unit 104, a plan for achieving the above-mentioned purpose including a plurality of topics as shown in FIG.
[0029]
Each topic corresponds to a problem to be solved to achieve the above purpose. Each task has a condition for solving the problem (a condition that should be satisfied for determining that the problem has been solved), an operation for satisfying the condition, and an operation for satisfying all conditions ( The operation when all conditions are achieved) and the output content output as task output information when solving the problem are predetermined. That is, each task is associated with at least one condition for solving the problem, an operation for satisfying each condition, an operation when all conditions are achieved, and a task including output contents when the problem is solved. Have been. Here, a description of a condition included in a task corresponding to each task, an operation, an operation when all conditions are achieved, and an output content when the task is solved is referred to as task information.
[0030]
A main task for “destination setting” is described in the task information T1.
[0031]
The task represented by the task information T1 is "destination setting" as the purpose of the dialog (described in the "name" field). As a condition for solving this problem, it is necessary to obtain a “destination approved place” (described in the “condition” field), and as an operation for that, a process described in the task information T2 is performed. Performs the operation (solves the problem described in the task information T2), and then transitions to the processing operation of the task information T3, and performs the processing operation described in the task information T3 (the processing described in the task information T3). Solve the problem). As a result, when the above condition is satisfied, the operation described in the “operation when all conditions are achieved” field (here, “calculate the route to the designated place”) is performed, and the result is output to the task output information. Output.
[0032]
In the following description, a processing operation based on the description of task information is simply called a task. For example, a processing operation based on the description of the task information T2 is called a task T2.
[0033]
The task information T2 is a processing operation for “location search”. In order to solve this problem, first, task information T5 for which “search method designation” is to be solved is called, and The described processing operation is performed.
[0034]
The processing operation described in the task information T5 utters “Which address or registered name do you want to search for?” (Described in the “operation” field), and the utterance input by the user accordingly. The search method specified by the user is obtained from the content (described in the “condition” field), and the search corresponding to the specified search method is described as described in the “operation when all conditions are achieved” field. Call task information to do (load task). Here, the task information T6 is read when the search method by the address is specified, and the task information T7 is read when the search method by the registered name is specified.
[0035]
The processing operation described in the task information T6 has a task of “searching for a place by an address”. The user inputs an address as described in the “operation” fields corresponding to each of the three conditions. When the user speaks for the address, and the address is obtained (described in the "condition" field), as described in the "action when all conditions are achieved" field, the "address specified" Search for corresponding location "and output the search result as task output information.
[0036]
The processing operation described in the task information T7 has a task of “searching for a place by a registered name”, and as described in the “operation” field, gives an utterance for the user to input the registered name. As a result, when the registered name is obtained (described in the “condition” field), as described in the “operation when all conditions are achieved” field, the “location corresponding to the specified registered name is determined. Search ”and output the search result as task output information.
[0037]
When the search result is notified as task output information by the task T6 or the task T7, the condition “searched place” described in the “condition” field of the task information T4 is satisfied. Since the task information T4 does not describe "operation when all conditions are achieved", in this case, the process returns to the task T2 that called (loaded) the task T4. From the task T4, the “searched place” output from the task T6 or the task T7 is output as the task output information “search result place”, and this is information to be notified to the task T2.
[0038]
In the task T2, since the condition of “acquire a place” described in the “condition” field is satisfied by the input information, the operation described in the “operation when all conditions are achieved” field, that is, transition to the task T3 is performed. At the same time, the “location of the search result” is output, and this is input information to the task T3.
[0039]
In task T3, the task is to obtain approval from the user for the "location of search result" as the input information. Therefore, an operation for that is performed, and if approval is obtained, an approved flag is given to the location. Then, a place approved as a destination is output as task output information.
[0040]
Since the approved place is obtained by the above processing operation (since the condition of the “condition” field of the task information T1 is satisfied), as described in the “operation when all conditions are achieved” field of the task information T1. Calculate the route to the place specified as the purpose, and output the result as task output information.
[0041]
As described above, when one task includes another task hierarchically (nested) (corresponding to the relationship between tasks T1 and T2, tasks T2 and T4, task T4 and task T5, etc.), When reading out a task, the task is called to load the other task. When one task includes a plurality of tasks in parallel (for example, the relationship between task T2 and task T3 in task T1), transition to task T3 after task T2 ends is referred to as transition.
[0042]
Here, in the description of each task information, the output content when all the conditions are achieved is described in the “output” field. This output content corresponds to the task output information.
[0043]
In the “condition” field, a condition for solving the task of the task (or a condition to be satisfied for determining that the task is solved) is described, and “operation” corresponding to each “condition” field is described. In the field, a processing operation (procedure) required to satisfy the condition is described. To describe progressing to the next task after achieving a certain condition, another task is nested inside the “operation” field as in the task information T1.
[0044]
The candidate selection unit 101 of the dialogue device 100 selects a current dialogue state or a past dialogue based on the dialogue information stored in the dialogue information storage unit 103 from among the plurality of recognition candidates output from the natural language interpretation device 1. An optimum one is selected from the situation and is output to the dialog stack operation unit 102 as input information.
[0045]
The dialogue information storage unit 103 stores dialogue information for managing the dialogue status up to the present according to the input information input from the candidate selection unit 101. Here, in order to manage the dialogue status, the dialogue information storage unit 103 stores the dialogue information corresponding to each of the unsolved problem and the current (unsolved) problem used in the past as the purpose of the dialogue with the user. .
[0046]
Each piece of conversation information includes a task (task information) used in the past or the present as a conversation purpose, a focused condition (attention condition) among conditions in the task information, and intention information.
[0047]
The conversation information storage unit 103 is a stack that stores conversation information in a LIFO (Last In First Out) method.
[0048]
The dialog stack operation unit 102 stores (push) the dialog information corresponding to the task in the dialog information storage unit 103 (stack) in the order of the tasks used for the purpose of the dialog. The task corresponding to the last stored conversation information (that is, the conversation information stored at the top of the stack) among the conversation information stored in the stack is the current conversation purpose, and the problem as the conversation purpose is solved. Then, the conversation information is extracted from the stack (pop).
[0049]
The conversation stack operation unit 102 manages conversation information by performing a stack operation (push, pop). That is, the conversation status at a certain point is described in the conversation information (stack top) indicated by the top address of the stack at that time.
[0050]
For example, suppose that the dialogue system shown in FIG. 1 was started, and the user made a statement “set a destination”. At this time, it is assumed that “destination setting” is selected by the candidate selection unit 101 of the dialog device 100 as a result of the interpretation by the natural language interpretation device 1 and is input to the dialog stack operation unit 102 as input information. Since the task corresponding to (related to) the input information is the task T1 (since the task (name) of the task information is “destination setting”), the task (task T1) Is generated in correspondence with the input information “destination setting” (see FIG. 6) and pushed (written) on the stack. Then, the task T1 is executed. In this case, the condition “destination approved place” for solving the task of the task T1 has not been obtained yet. An operation for satisfying the condition, that is, the task information T2 is loaded, and a condition “acquire a place” for solving the task of the task T2 is set as a notice condition of the dialog information. FIG. 7 shows the conversation information at this time.
[0051]
Note that the task T2 is loaded from the task T1, that is, since the task T2 is in a relationship nested (nested) in the task T1, as shown in FIG. The "destination location setting" and the condition "acquire location" are represented by connecting the name and condition of the task T1 with a hyphen "-".
[0052]
As described above, when it is necessary to solve another problem (issue y) in order to satisfy a certain condition for solving a certain problem (issue x), it can be said that the problem y is in a relationship included in the problem x. However, in this case, the condition for solving the problem y can also be said to be the condition for solving the problem x.
[0053]
When there is an inclusion relationship between the task x and the task y, the dialog information corresponding to the task y is described by connecting the name and condition of the dialog information corresponding to the task x with a hyphen, and the dialog information corresponding to the task x May be included.
[0054]
By the way, when the stack is in the state shown in FIG. 7, an operation for satisfying the condition described in the task information T2 is executed. In this case, the task T2 is loaded. When the task T2 is loaded, new interaction information E2 is generated by associating the task T2 with the name "location search" of the name of the task T2 as intention information, and is pushed onto the stack (see FIG. 8).
[0055]
The intention information of the dialog information means information that mainly represents the user's intention obtained through the dialog with the user when the task corresponding to the dialog information is intended for the dialog. In this embodiment, as described later, the task output information is handled in the same manner as the user's intention, and is associated with the dialog information as the intention information.
[0056]
When the dialogue proceeds according to the plan (for example, without interruption), as described above, the dialogue information is pushed down to the stack, and the processing operation of the task corresponding to each piece of the dialogue information ends (the task of the task is determined as follows). To be solved) and pop-up (take-out) in order from the conversation information at the top of the stack, and when all the conversation information is gone, the purpose of the conversation is achieved.
[0057]
In the interactive device according to the present embodiment, when input information that is one of utterance information or task output information representing the utterance content of the user is input, the input information and all the interactive information in the stack are input. Is compared with the input information, that is, the dialog information corresponding to the task in which at least one condition is satisfied with the input information in the current and past dialog information, and the same as the input information. A search is performed to determine whether or not there is dialog information in which the type of information is recorded as intention information. If such dialog information exists, the dialog information is set as a stack top, and a task corresponding to the dialog information is set as a dialog purpose. .
[0058]
When all the conditions of the task as the conversation purpose are satisfied, the conversation information corresponding to the task is deleted from the stack, and information output from the task of the task (output content described in the “output” field) And all the conversation information in the stack, and in the past conversation information, conversation information corresponding to a task that satisfies at least one condition in the output content of the task, the same type as the output content Is searched for dialog information in which the information of the above is recorded as intention information. When there is such conversation information, a stack operation is performed so that the conversation information becomes the top of the stack. It should be noted that when the problem corresponding to the newly-topped dialogue information satisfies all the conditions by the output contents, and the problem is solved, the dialogue information is also extracted from the stack.
[0059]
The dialogue purpose is a problem corresponding to the dialogue information at the top of the stack. More specifically, a condition which is set as a noticeable condition of the dialogue information and which must be satisfied in order to solve the problem (such as When there are a plurality of conditions, one of them).
[0060]
The conversation stack operation unit 102 executes load, transition, clear, and selection of an achievement condition, in addition to push / pop, as a stack operation.
[0061]
The load is a procedure (operation) used when calling a new task, and pushes conversation information corresponding to the new task onto a stack. This is an operation for calling the task of the task y when it is necessary to solve another task (the task y) in order to satisfy a certain condition for solving the certain task (the task x).
[0062]
The transition is also a procedure for calling a new task, but is a procedure for calling a task y when solving a task in order and moving to another task y when a certain task x is solved. .
[0063]
In order to clear a task x as a conversation purpose, clearing removes the conversation information stored in the stack after the conversation information corresponding to the task x is stored in the stack, and stores the conversation information corresponding to the task x in the stack top. The procedure is as follows.
[0064]
The achievement condition selection is a procedure for selecting another condition for solving the task, such as when the attention condition of the task corresponding to the conversation information at the top of the stack is satisfied. If another task (task) is nested in the "action" field in the task information corresponding to the selected condition, the achievement condition selection is recursively advanced for that task, and Write into the "attention condition" field in the dialog information.
[0065]
As an operation for satisfying a certain condition, as in task T5, when the dialogue system utters “Which address or registered name is to be searched?”, The response information representing the utterance content is a response. Output to the output device 2.
[0066]
The response output device 2 outputs the response information output from the dialog stack operation unit 102 to the user in a predetermined format, for example, in the form of voice, text, or the like. Since the details of the response output device 2 are not the gist of the present invention, the description thereof is omitted. However, in the present embodiment, for the sake of simplicity, the utterance content described as the operation in the task information is directly output as voice or It shall be output as text. In addition to the above, an existing one can be used, such as outputting a sentence using a template from intention information.
[0067]
Next, the processing operation of the dialog stack operation unit 102 of the dialog system of FIG. 1 will be described with reference to flowcharts shown in FIGS.
[0068]
FIG. 3 is a flowchart for explaining the entire processing operation of the dialog stack operation unit 102. This processing operation is performed by using information obtained from a user's statement from the candidate selection unit 101 to the dialog stack operation unit 102 (hereinafter, the statement Is called when (information) is input (step S0). FIGS. 4 and 5 are flowcharts for explaining the notification destination search processing operation in step S2 in FIG. 3, and include the utterance information input in step S0 in FIG. 7 shows a processing operation of searching for a task as a notification destination of information (task output information) output from the task of the subject based on the conversation information in the stack. Here, the task as the notification destination is one of the tasks stored as the conversation information in the stack. For example, the utterance information or the task output information determines that the task has been solved. The task is a task corresponding to a task that satisfies at least one of one or more conditions for the task, or another task for solving the task.
[0069]
First, the processing operation of the dialog stack operation unit 102 will be described with reference to FIG.
[0070]
Here, the buffer BF1 as a buffer for temporarily storing the input utterance information and the task output information output from the task, and the buffer currently stored in the buffer BF1 are the utterance information input from the candidate selection unit 101. At some point, a new flag that is set to "1" is used.
[0071]
First, when the utterance information is input in step S0, the utterance information is stored in the buffer BF1, and "1" is set in the new flag (step S1). Next, the process proceeds to step S2, and a task as a notification destination of the information stored in the buffer BF1 is searched based on the conversation information in the stack. The details will be described later.
[0072]
When a task as a notification destination is found in step S2, the process proceeds to step S4, and when there is no task as a notification destination, the process proceeds to step S13 (step S3).
[0073]
In step S4, a clear operation is performed so that the conversation information corresponding to the task as the notification destination is at the top of the stack, and the statement information stored in the buffer BF1 is reported in the "intention information" field of the conversation information. I do. In this case, "notification" refers to an operation of overwriting information corresponding to the same condition in a task and adding information corresponding to a different condition.
[0074]
Then, it is checked whether or not the information currently recorded in the “intention information” field of the dialog information, that is, the intention information satisfies all the conditions for solving the problem corresponding to the dialog information ( Step S5). When the intention information satisfies all the conditions for solving the problem, the process proceeds to step S6, and when there is at least one condition that the intention information does not satisfy, the process proceeds to step S9.
[0075]
In step S6, the procedure specified in the "operation when all conditions are achieved" field of the task corresponding to the conversation information at the top of the stack (the task corresponding to the task in which all conditions have been achieved and solved) is executed. Then, the task output information is stored in the buffer BF1 according to the description of the "output" field of the task (step S7a). Since the task output information is currently stored in the buffer BF1, the new flag is cleared to “0” in step S7 (step S7b).
[0076]
Further, since the task of the task has been achieved (since there is no need to continue the dialogue to solve the task), the conversation information is deleted from the stack. At this time, if the conversation information is not included in other conversation information, the conversation information (also called a stack element) is popped from the stack.
[0077]
Next, the process proceeds to step S8, and it is confirmed whether or not another task is being loaded and transitioned during the execution of the operation when all conditions are satisfied in step S6. If all tasks are loaded and transitioned as tasks when all conditions are achieved (other tasks are loaded and transitions are specified in the task information), the dialog corresponding to the loaded or transitioned task is performed. Information is stored at the top of the stack. Then, the process returns to step S4 in order to set the task corresponding to the conversation information as the notification destination. On the other hand, if no other task is loaded or transitioned as an operation when all conditions are achieved, the process returns to step S2 to search for a task as a notification destination of the task output information stored in the buffer BF1.
[0078]
In step S9, an achievement condition selection operation is performed on the stack top. That is, another condition for solving the problem corresponding to the conversation information at the top of the stack is selected, and the selected condition is written in the “attention condition” field as the attention condition. Then, the process proceeds to step S10 to execute an operation for obtaining information that satisfies the current attention condition. In this operation, when another task is loaded (step S11), the conversation information corresponding to the loaded or transitioned task is stored in the stack top. Therefore, the name of the task is stored in the buffer BF1 (step S12a), and the task of the notification destination is set to a new flag “0” in order to set the task corresponding to the dialog information to the notification destination (step S12b). ), And return to step S4.
[0079]
At the time of executing the operation corresponding to the attention condition, if the loading or transition of another task is not specified in the task information (step S11), the process ends here.
[0080]
In step S3, when the task as the notification destination of the utterance information or the task output information stored in the buffer BF1 cannot be obtained from the conversation information stored in the stack up to the present, the process proceeds to step S13.
[0081]
If the new flag is "1" in step S13, since what is currently stored in the buffer BF1 is the utterance information obtained from the utterance content from the user, the process proceeds to step S14, where the utterance information is stored. A corresponding new task is obtained (for example, a task having a name that matches the utterance information is loaded), and conversation information corresponding to the new task is stored in the stack. On the other hand, when the new flag is “0” in step S13, what is currently stored in the buffer BF1 is the task output information, which is used to determine the problem solved by the task output information and the fact that the task has been solved. Since there is no problem in the stack that satisfies at least one of the plurality of conditions, the process ends here.
[0082]
At this time, if the conversation information remains in the stack, the operation corresponding to the attention condition of the conversation information at the top of the stack may be executed again to solve the remaining conversation information.
[0083]
Next, the notification destination search processing operation in step S2 in FIG. 3 will be described with reference to FIGS.
[0084]
Here, a buffer BF2 is used to store the conversation information in the current stack to be searched, and the conversation information stored in the buffer BF2 is referred to as a “target stack”. In addition, a buffer BF3 is used to store task information to be searched at present, and the task information (task) stored in the buffer BF3 is referred to as a “target task”. Both are empty (NULL) at the initial stage (step S100).
[0085]
When the notification destination search process is started (when the search for the notification destination of the utterance information or the task output information stored in the buffer BF1 is started), the buffer BF1 sequentially starts from the conversation information at the top of the stack (from the top of the stack). The dialog information stored in the lower order is stored one by one (step S102), and the following notification destination search processing is repeated until the search for all the dialog information in the stack is completed (step S102). Step S101).
[0086]
First, in step S103, the conversation information at the top of the stack is stored in the buffer BF2, and the task information corresponding to the conversation information set in the “target stack” is stored in the buffer BF3 (set as the “target task”). .
[0087]
The utterance information or task output information stored in the buffer BF1 is referred to as input information in the notification destination search processing.
[0088]
First, the process proceeds to step S104 in FIG. 5, and it is checked whether the input information satisfies the target condition of the “target stack”. If the condition is satisfied, the process proceeds to step S108, and the current “target task” is output as the task to be notified. If the condition is not satisfied, the process proceeds to step S105.
[0089]
In step S105, it is checked whether the input information satisfies all the conditions of the “target task”. If it satisfies, the process proceeds to step S108, and outputs the current “target task” as the task to be notified. If it does not, the process proceeds to step S106.
[0090]
In step S <b> 106, it is determined whether or not there is a task whose input information satisfies all conditions among the tasks nested in the operation of the target condition of the “target stack”. When there is such a task, the process proceeds to step S107, and the task is set as a “target task”. When there is no such task, the process proceeds to step S109.
[0091]
Here, in step S104 and step S106, when the target condition of the “target stack” is the condition of the nested task, the condition defined in the “target task” is selected. For example, in FIG. 7, the attention condition is “accepted place−acquired place” of the nested task T2, but if the “target task” is T1, the “accepted place” is adopted as the attention condition. I do.
[0092]
In step S109, the input information is compared with the intention information described in the “intention information” field of the “target stack”, and it is checked whether or not they are the same type of information. If the information is of the same type, the process proceeds to step S108, and the current “target task” is output as the task to be notified. If the information is different, the process proceeds to step S110.
[0093]
In step S110, it is checked whether there is any task that includes the current “target task” in a nested manner. If there is a task including the “target task”, the process proceeds to step S111, the task is set as the “target task”, and the process returns to step S104. On the other hand, when there is no task including the “target task”, the process returns to step S101, and the notification destination search process is repeated with the next conversation information in the stack as the target stack.
[0094]
As a result of searching for a notification destination as described above based on all the conversation information in the stack, if a task as a notification destination is not obtained (step S101), the process proceeds to step S112, and “no notification destination” is output. I do.
[0095]
The notification destination search processing is based on the premise that "if the user is cooperative, a response related to the current conversation is made", and the input information can be reported retrospectively from the top of the stack, which is the current conversation contents. Stacks and tasks are to be searched.
[0096]
Subsequently, the processing operation of the candidate selection unit 101 will be described. The candidate selection unit 101 also operates based on the notification destination search processing of FIGS. From the group of utterance information input from the natural language interpreting apparatus 1, the likelihood is set so as to give priority to a candidate that can be notified to the top of the stack and, in the case of the same stack element, can be notified to the one closer to the currently focused dialogue Edit (score).
[0097]
More specifically, in order to notify the top of the stack (the task corresponding to the dialog information at the top of the stack), it is necessary to reduce the number of passes in step S101 in FIG. In order to notify the candidate, priority is given to a candidate that reduces the number of passes in steps S104, S105, S106, and S109 in FIG.
[0098]
For example, the operation may be performed so as to lower the score of the candidate each time the candidate passes through steps S101, S104, S105, S106, and S109. Thereby, it is possible to preferentially select an utterance having contents suitable for the dialog at that time.
[0099]
Here, the processing operation of the dialogue device will be described more specifically taking a dialogue system applied to a car navigation system as an example.
[0100]
It is assumed that the user's intentions input to the car navigation system are, for example, four types: destination setting, location search, location search by address, and location search by registered name. These can be said to be a plurality of problems to be solved for setting a destination. FIG. 2 shows task information corresponding to each task.
[0101]
The stack in the initial state is assumed to be empty. Here, for the sake of simplicity, it is assumed that information obtained as a recognition result of a user's utterance input to the interactive device 100 has no ambiguity, and the result of interpretation by the natural language interpreting device 1 is directly used as utterance information in the interactive stack operation unit 102. The detailed operation of the candidate selection unit 101 will be described later.
[0102]
(1) First remark: It is assumed that “set a destination” is input to the natural language interpreting apparatus 1.
[0103]
It is assumed that the interpretation result of the natural language interpretation device 1 is “destination setting”. When this is notified to the dialog stack operation unit 102 as utterance information, the following operation is performed according to the flowchart of FIG.
[0104]
First, "destination setting" is stored in the buffer BF1, and the new flag is set to "1" (step S0, step S1). In step S2, the notification destination is searched. However, since the stack is empty, it is determined that there is no notification destination (steps S101 and S112 in FIG. 4), and since the new flag is "1", the process proceeds to step S14. The task corresponding to the utterance information “destination setting”, that is, the task T1 having the name of the utterance information “destination setting” is selected from the plan storage unit 104 and loaded. At this time, the dialog stack operation unit 102 generates the dialog information E1 (see FIG. 6) corresponding to the task T1, and pushes it to the stack.
[0105]
Since the task T1 is called in response to the comment information “destination setting”, the current conversation purpose of the user is “destination setting”, which is written in the conversation information E1 as intention information as shown in FIG. (Step S4).
[0106]
At this time, the stack is in a state where only the conversation information E1 is stored as shown in FIG.
[0107]
The task T1 has only one condition and is “approved place”, but in the state of the stack in FIG. 6, information satisfying this condition has not been obtained yet (step S5). Therefore, “approved place” is selected as the attention condition (step S9). Since the operation corresponding to the condition “approved place” includes the task T2 “destination place setting” (nested), the condition of the task T2, that is, “acquire the place” Is also selected. As a result, as shown in FIG. 7, the dialogue information E1 has the name “destination setting—destination location setting” and the attention condition “accepted location—acquisition of location,” as shown in FIG. (Step S9).
[0108]
Subsequently, an operation corresponding to the attention condition “acquire a place” set in the interaction information E1 is executed (step S10). In this operation, the task T4 of “location search” is loaded. Therefore, the task T4 is loaded by this designation (step S11), and the call to the buffer BF1 is made according to the user's statement. Instead, the name “location search” of the task T4 is stored, and the new flag is set to “0” (step S12a, step S12b).
[0109]
As shown in FIG. 8, the task T4 is loaded, and the conversation information E2 corresponding to the task T4 is pushed onto the stack (the conversation information E2 is stacked on the conversation information E1). Since the task T4 is notified of "location search" stored in the buffer BF1, "location search" is written as the intention information of the dialog information E2 (step S4). As shown in FIG. 8, the conversation information E2 is currently at the top of the stack, and the task corresponding to the task T4 is the current conversation purpose.
[0110]
The condition of the task T4 is only one “searched place”, but information satisfying this condition has not been obtained yet in the state of the stack in FIG. 8 (step S5). Therefore, “searched place” is selected as the attention condition (step S9). Since the operation corresponding to the condition “searched place” includes the task T5 “designation of search method” (nested), the condition of this task T5, that is, “acquire search method” Is also selected. As a result, as shown in FIG. 9, the name of the conversation information E2 is “place search—designation of search method”, and the attention condition is “acquired place—acquire search method”, as shown in FIG. (Step S9).
[0111]
Subsequently, an operation corresponding to the attention condition “acquire search method” set in the interaction information E2 is executed (step S10). Since this operation is to present a sentence to the user, response information indicating the content is output to the response output device 2. At this point, the processing shown in FIG. 3 ends. FIG. 9 shows the state of the stack at this time.
[0112]
The response output device 2 outputs a response sentence "Which address or registered name do you want to search?" As a result, when there is a utterance from the user (when utterance information obtained from the content of the utterance of the user is input), the processing is started again from step S0 in FIG.
[0113]
(2) Second remark: It is assumed that "search by address" is input to the natural language interpreting apparatus 1.
[0114]
It is assumed that the interpretation result of the natural language interpretation device 1 is “location search by address”. When this is notified to the dialog stack operation unit 102, it operates as follows according to the flowchart of FIG.
[0115]
First, "location search by address" is stored in the buffer BF1, and the new flag is set to "1" (step S0, step S1). In step S2, a notification destination is searched. Here, since the statement information satisfies the attention condition of the dialog information E2 in FIG. 9 (step S104 in FIG. 5), the task to which the statement information is notified is the task T5. Therefore, as shown in FIG. 10, “location search by address” is further added to the intention information of the conversation information E2 corresponding to the task T5 (step S4).
[0116]
As a result, the task T5 corresponding to E2 in the conversation information at the top of the stack satisfies all necessary conditions (step S5), and executes the operation at the time of achieving all the conditions predetermined for the task T5 (step S5). Step S6). Since the task of task T5 has been achieved, the conversation information corresponding to the nested task T5 is deleted from the stack. For example, in this case, “designation of search method” and “acquisition of search method” are deleted from the name and attention condition of the interaction information E2.
[0117]
Since the corresponding search is loaded in the operation at the time of achieving all the conditions, the task T6 of “location search by address” is loaded by this designation (step S11), and the task T6 is loaded into the buffer BF1. The intention of calling is that the name is “search by address” in the task T6, and the new flag is set to “0” (step S12a, step S12b).
[0118]
As shown in FIG. 11, the task T6 is loaded, and the conversation information E3 corresponding to the task T6 is pushed onto the stack (the conversation information E3 is stacked on the conversation information E2). Since the task T6 is notified of the "location search by address" stored in the buffer BF1, "location search by address" is written as the intention information of the dialogue information E3 (step S4). As shown in FIG. 11, the conversation information E3 is currently the top of the stack, and the task corresponding to the task T6 is the current conversation purpose.
[0119]
Thereafter, the process proceeds in the same manner as described above, and in step S9, if "obtain prefecture name" is selected as the attention condition, in step S10, the operation corresponding to the attention condition is "Please enter the prefecture name." Is presented to the user. Since there is no operation to load another task in this operation, the process shown in FIG. 3 is temporarily terminated at this point. FIG. 12 shows the state of the stack at this point.
[0120]
(3) Third remark: "Kanagawa prefecture" is input to the natural language interpreting apparatus 1.
[0121]
It is assumed that the interpretation result of the natural language interpreting apparatus 1 is “location search by address: Kanagawa” including intention information “location search by address” of the current stack top dialogue information. When this is notified to the dialog stack operation unit 102, it operates as follows according to the flowchart of FIG.
[0122]
First, in the buffer BF1, utterance information "location search by address: Kanagawa" is stored, and a new flag is set to "1" (step S0, step S1). In step S2, a notification destination is searched. Here, since the statement information satisfies the attention condition of the conversation information E3 at the top of the stack in FIG. 12 (step S104 in FIG. 5), the task to which the conversation information is notified is the task T6. Accordingly, "location search by address: Kanagawa" is written in the intention information of the dialog information E3 corresponding to the task T6 (step S4). Since all the conditions for achieving the task of task T6 have not been satisfied yet (step S5), "obtain municipalities" is selected as the attention condition from the remaining conditions (step S9). By executing an operation predetermined for the attention condition, the response output device 2 outputs “please enter the city, town, and village name”. FIG. 13 shows the state of the stack at this time. Since there is no operation to load another task in this operation, the process shown in FIG. 3 is temporarily terminated at this point.
[0123]
(4) Fourth remark: It is assumed that “after all, with the registered name” is input to the natural language interpreting apparatus 1.
[0124]
It is assumed that the interpretation result of the natural language interpreting apparatus 1 is “location search by registered name” based on the current stack-top conversation information. When this is notified to the dialog stack operation unit 102, it operates as follows according to the flowchart of FIG.
[0125]
First, the utterance information “location search by registered name” is stored in the buffer BF1, and the new flag is set to “1” (step S0, step S1). In step S2, a notification destination is searched. However, task information and intention information satisfying at least one of all conditions or a plurality of conditions for solving the problem are obtained by the statement information “location search by registered name”. The same type of conversation information does not exist (step S104 to step S106, step S109). Therefore, there is no task as a notification destination in the stack. In this case, since the new flag is “1” (step S13), the task corresponding to the statement information, that is, the task T7 having the name of the statement information “location search by registered name” is stored in the plan storage unit 104. Select and load. At this time, the dialog stack operation unit 102 generates the dialog information E4 (see FIG. 14) corresponding to the task T7, and the dialog information E4 is pushed (the dialog information E4 is stacked on the dialog information E3). .
[0126]
Since the task T7 is called upon receiving the comment information “location search by registered name”, the current user's dialogue purpose is “location search by registered name”, which is the dialogue as shown in FIG. The information is written in the information E4 (step S4).
[0127]
FIG. 14 shows the stack at this time.
[0128]
As the attention condition, only one condition “acquire registered name designation” of task T7 is selected, and by executing an operation corresponding to the attention condition, “please enter registered name” is presented to the user. FIG. 15 shows the state of the stack at this time. The task corresponding to the conversation information at the top of the stack is a location search using the registered name of task T7.
[0129]
(5) Fifth remark: It is assumed that "Minato Mirai" is input to the natural language interpreting apparatus 1.
[0130]
It is assumed that the interpretation result of the natural language interpreting apparatus 1 is “location search by registered name: Minato Mirai” including the intention information “location search by registered name” of the current stack top conversation information. When this is notified to the dialog stack operation unit 102, it operates as follows according to the flowchart of FIG.
[0131]
First, in the buffer BF1, utterance information "location search by registered name: Minato Mirai" is stored, and the new flag is set to "1" (step S0, step S1). In step S2, a notification destination is searched. Here, since the statement information satisfies the attention condition of the conversation information E4 at the top of the stack in FIG. 15 (step S104 in FIG. 5), the task to which the conversation information is notified is the task T7. Therefore, “location search by registered name: Minato Mirai” is written in the intention information of the conversation information E4 corresponding to the task T7 (step S4). Since all the conditions for achieving the task of the task T7 are satisfied by this intention information, that is, since the registered name has been obtained (step S5), the operation at the time of achieving all the conditions predetermined for the task T7 Is executed (step S6). When the operation at the time of achieving all the conditions is executed, a place having a registered name of “Minatomirai” (here, this is represented as “minatomirai”) is searched. Since the task output information predetermined for the task T7 is “location of search result”, “minatomirai” which is the search result of the operation when all the conditions are achieved is stored in the buffer BF1 as task output information, and a new flag is set. Is set to "0" (step S7a, step S7b).
[0132]
Since the task of the task T7 has been achieved, the conversation information corresponding to the task T7 is deleted from the stack. For example, in this case, the conversation information E4 is deleted. FIG. 16 shows the state of the stack at this time. That is, the current stack top is the conversation information E3.
[0133]
Next, in step S2, the notification destination of the task output information “minatomirai” currently stored in the buffer BF1 is searched.
[0134]
Since the task output information is information that satisfies the attention condition “searched place” of the conversation information E2 in the stack shown in FIG. 16 (step S104 in FIG. 5), the notification destination corresponds to the conversation information E2. It becomes task T4 (step S3).
[0135]
In this case, a clear operation is performed so that the conversation information E2 is at the top of the stack. That is, the dialog information E3 is popped from the stack. Then, the task output information stored in the buffer BF1 is notified to the task T4 corresponding to the dialog information E2, and thus "minatomirai" is added to the intention information of the dialog information E2. FIG. 17 shows the state of the stack at this time.
[0136]
When the task output information stored in the buffer BF1 is notified, all the conditions of the task T4 are satisfied (step S5). Since the task T4 does not have an operation when all the conditions are achieved, the task output information predetermined for the task T4, that is, “minatomirai” is stored in the buffer BF1 as the “location of the search result”, and the new flag is set. "0" is set (steps S7a, 7b). Since the task corresponding to the task T4 has been achieved, the dialog information corresponding to the task T4, that is, the dialog information E2 is also popped. The stack top is the conversation information E1.
[0137]
Next, in step S2, a notification destination of the task output information currently stored in the buffer BF1 is searched. Since the task output information is information that satisfies the attention condition “acquire a place” of the conversation information E1 in the stack shown in FIG. 18 (step S104 in FIG. 5), the notification destination is the conversation information E1 on the stack top. (Step S3).
[0138]
Since the task output information stored in the buffer BF1 is notified to the task T2 corresponding to the conversation information E1, "minatomirai" is added to the intention information of the conversation information E1.
[0139]
When the task output information stored in the buffer BF1 is notified, all the conditions of the task T2 are satisfied (step S5). Therefore, the dialog information of the task T2 included in the task T1 is deleted from the dialog information E1. For example, in this case, “destination location” and “acquire location” are deleted from the name and attention condition of the conversation information E1. FIG. 19 shows the state of the stack at this time.
[0140]
Since the transition to the task T3 is determined in advance in the task T2 as the operation when all conditions are achieved, the task T2 transitions to the task T3 (step S6). Then, the task output information predetermined in the task T4, that is, “minatomirai” is stored in the buffer BF1 as the “location of the search result”, and the new flag is set to “0” (steps S7a and 7b). ).
[0141]
Since the task T3 is a task included in the task T1, when transitioning to the task T3 (step S8), "destination approval" which is the name of the task T3 is added to the name of the conversation information E1 (step S8). S4). Since the condition "consent to the searched place" for achieving the task of task T3, that is, the user's consent to the searched place has not been obtained yet (step S5), the attention condition of the conversation information E1 includes the task The condition of T3, "Consent to the searched place" is set (step S9). Then, an operation corresponding to the attention condition is executed (step S10). By executing this operation, "Do you want to set this as a destination?" Is output to the user, and the position of "minatomirai" is displayed on the map. Since there is no load processing in the "operation" field, the processing is terminated here and the input from the user is waited. FIG. 20 shows the state of the stack at this point.
[0142]
(6) Sixth remark: It is assumed that "yes" is input to the natural language interpreting apparatus 1.
[0143]
It is assumed that the interpretation result of the natural language interpretation device 2 is “permitted”. When this is notified to the dialog stack operation unit 102, it operates as follows according to the flowchart of FIG.
[0144]
First, the utterance information "permission" is stored in the buffer BF1, and the new flag is set to "1" (step S0, step S1). In step S2, a notification destination is searched. Here, since the comment information satisfies the attention condition of the conversation information E1 at the top of the stack in FIG. 20 (step S104 in FIG. 5), the task to which the conversation information is notified is the task T3. Therefore, the statement information “consent” stored in the buffer BF1 is written in the intention information of the conversation information E1 corresponding to the task T3 (step S4). Since all the conditions for achieving the task of the task T3 are satisfied by this intention information, that is, the consent to the searched place has been obtained (step S5), all of the predetermined tasks T3 have been set. The operation at the time of condition achievement is executed (step S6). Executing this all-conditions-fulfilled action will give the accepted location an accepted flag (designated "destOK"). Since the task output information predetermined for the task T3 is the “location of the search result”, the accepted location provided with the accepted flag, which is the result of the operation when all the conditions are achieved, that is, “minatomirai ( "destOK)" is stored in the buffer BF1 as task output information, and the new flag is set to "0" (steps S7a and S7b). Since the task of the task T3 has been achieved, the conversation information corresponding to the task T3 is deleted from the stack. For example, in this case, “destination approval” and “acceptance for the searched place” are deleted from the name and attention information of the conversation information E1. In the operation at the time of achieving all the conditions, since there is no load process, the process proceeds to step S2. FIG. 21 shows the state of the stack at this time.
[0145]
The information currently stored in the buffer BF1 is the accepted location given the accepted flag, which is the task output information of the task T3, that is, “minatomirai (destOK)”. Since this task output information is information that satisfies the attention condition “destination approved place” of the conversation information E1 in the stack shown in FIG. 21 (step S104 in FIG. 5), the notification destination is the conversation information E1. The task becomes the corresponding task T1 (step S3). Since the task output information stored in the buffer BF1 is notified to the task T1 corresponding to the dialog information E1, "minatomirai (destOK)" is added to the intention information of the dialog information E1. FIG. 22 shows the state of the stack at this time.
[0146]
When the task output information stored in the buffer BF1 is notified, all the conditions of the task T1 are satisfied (step S5), and the operation at the time of achieving all the conditions predetermined for the task T1 is executed (step S6). ). When the operation at the time of achieving all the conditions is executed, the route calculation to the accepted location “minatomirai” is performed. Since the task output information predetermined for the task T1 is “route information”, the route information up to “minatomirai”, which is the calculation result of the operation when all conditions are achieved, is stored in the buffer BF1 as task output information. The new flag is set to "0" (step S7a, step S7b). Since the task of the task T1 has been achieved, the conversation information E1 corresponding to the task T1 is popped from the stack. In this case, the stack is empty.
[0147]
Since the load operation is not performed in the operation when all the conditions of the task T1 are satisfied, the process proceeds to step S2. However, since the stack is empty (step S101 in FIG. 4), it is determined that there is no notification destination. Further, since the new flag is "0" (step S13), the input wait state for the user is entered.
[0148]
(7) In the state where the stack is as shown in FIG. 20, when "Do you want to set this as the destination?" Is output, the sixth remark: the user changes the place instead of "Yes". For this purpose, it is assumed that a seventh statement: "At Yamashita Park" is stated, and this is input to the natural language interpreting apparatus 1.
[0149]
It is assumed that the interpretation result of the natural language interpretation device 2 is “location search by registered name: Yamashita Park” by referring to, for example, the conditions of each task of the plan stored in the plan storage unit 104. When this is notified to the dialog stack operation unit 102, it operates as follows according to the flowchart of FIG.
[0150]
First, in the buffer BF1, utterance information "location search by registered name: Yamashita Park" is stored, and a new flag is set to "1" (step S0, step S1). In step S2, a notification destination is searched for. A task that satisfies at least one of all conditions or a plurality of conditions for solving the problem by using the statement information “location search by registered name: Yamashita Park” There is no dialog information of the same type as the intention information (step S104 to step S106, step S109). Therefore, there is no task as a notification destination in the stack. In this case, since the new flag is “1” (step S13), the task corresponding to the statement information, that is, the task T7 having the name “location search by registered name” included in the statement information is planned. Select from the storage unit 104 and load. At this time, the conversation stack operation unit 102 generates the conversation information E5 corresponding to the task T7, and the conversation information E5 is pushed (the conversation information E5 is stacked on the conversation information E1).
[0151]
Since the task T7 is called in response to the comment information “location search by registered name: Yamashita Park”, the current purpose of the dialogue of the user is “location search by registered name of Yamashita Park”. As shown in FIG. 23, the information is written in the conversation information E5 (step S4). FIG. 23 shows the stack at this time.
[0152]
Here, since the statement information written as the intention information in the conversation information E5 satisfies all the conditions for solving the task of the task T7, that is, since the registered name has been acquired (step S5), it is determined in advance in the task T7. The operation at the time of achieving all the conditions is executed (step S6). When the operation at the time of achieving all the conditions is executed, a location having a registered name of “Yamashita Park” (here, expressed as “yamashita-park”) is searched. Since the task output information predetermined for the task T7 is “location of search result”, “yamashita-park” which is the search result of the operation when all conditions are achieved is stored in the buffer BF1 as task output information. The new flag is set to "0" (step S7a, step S7b). Since the task of the task T7 has been achieved, the conversation information E5 corresponding to the task T7 is popped from the stack.
[0153]
Next, in step S2, a search is made for a notification destination of the task output information “yamashita-park” currently stored in the buffer BF1. In the process of step S2, when the "target task" is the task T3 corresponding to the dialogue information E1, the task output information does not indicate the permission of the user to satisfy the condition of the task T3, but has the same kind of intention. Since this is not the case (steps S104 to S106 and S109 in FIG. 5), the process proceeds to step S110 in FIG. The task T1 including the task T3 is referred to as a “target task”. The process proceeds to step S106 via steps S104 and S105. Here, all the conditions of the tasks T2 and T3 included in the operation related to the condition “approved place” corresponding to the task T1 in the attention condition of the conversation information E1 are checked.
[0154]
Then, since the task output information satisfies all the conditions “acquire a place” of the task T2, the task T2 is set as the “target task” (step S107), and the notification destination is determined to be the task T2 (step S108).
[0155]
In this case, the name of the dialogue information E1 is "destination setting-destination location setting", and the task output information stored in the buffer BF1 is notified to the task T2 corresponding to the dialogue information E1. Is rewritten to “yamashita-park” instead of “minatomirai”. When the task output information is notified to the task T2 corresponding to the dialog information E1 stored in the buffer BF1, all the conditions of the task T2 are satisfied (step S5). Therefore, the dialog information of the task T2 included in the task T1 is deleted from the dialog information E1. For example, in this case, "destination location" is deleted from the name of the conversation information E1. FIG. 24 shows the state of the stack at this time. At this stage, “yamashita-park” was obtained as the location of the search result.
[0156]
The state of the stack shown in FIG. 24 is the same as the state of the stack shown in FIG. In the same manner as described above, the subsequent processing operation transits to task T3, obtains consent for the location of the search result, returns to task T1, performs route calculation, and outputs the result.
[0157]
Next, the detailed operation of the candidate selection unit 101 will be described with reference to FIGS. 4 and 5, taking the case where the stack is in the state shown in FIG. 13 as an example. The score editing policy of the candidate selecting unit 101 in this example is such that the score is set to “−30” every time the process passes through step S101 and a lower-level stack element is searched, and steps S104, S105, S106, and S109 are performed. Is set to “−10” each time the vehicle passes, and the uniform score is set to “−50” by ignoring the deductions during the search when the notification destination is not found.
[0158]
When the stack is in the state shown in FIG. 13, it is assumed that the eighth statement: “Kawasaki City” is input to the natural language interpreting apparatus 1. This input is an example of the user's response to the response sentence “Please enter the city, town, and village name” output from the response output device 2. In response to this input, the natural language interpreting apparatus 1 obtains the following candidate groups by voice recognition.
"Permission (Yes)" (920)
"Kagawa Prefecture" (900)
"Kawasaki City" (880)
Note that “()” is a score given to the candidate (determined at the time of speech recognition).
[0159]
First, considering the candidate "permission", in step S103, the dialog information E4 is set for the "target stack", and the task T6 is set for the "target task". In any of step S104, step S105, step S106, and step S109, since the condition of the task "location search by address" is not satisfied by "permission", the process proceeds to step S110.
[0160]
Since the task including the task “location search by address” does not exist in the conversation information E3, the process proceeds to step S101, and the conversation information E2 is set in the “target stack”. Similarly, the task “location search” has no condition that is satisfied by “permission”, and the process proceeds to step S101 again to set the dialog information E1 in the “target stack”. In this case, “destination location setting” of the task T2 is initially set as the “target task” in step S103, but since there is no condition that is satisfied by “permission” in this task, the process proceeds to step S110. . Since the task T2 is included in the task T1, the task T1 “destination setting” is set as the “target task”, and the process proceeds to step S104. The process proceeds from step S104 to step S106 via step S105.
[0161]
In step S106, all the conditions of the tasks T2 and T3 included in the operation related to the attention condition “approved place” related to the task T1 are checked. Then, the task T3 “destination approval” is waiting for the user's permission, but is not satisfied because the resolved place has not been notified, and finally, the notification destination of “permission” cannot be found. Therefore, the score of the candidate “permitted” is “870”. Alternatively, if the task corresponding to “permission” is not defined, the new task cannot be loaded and is rejected.
[0162]
When the candidate “Kagawa Prefecture” is examined, in step S101, the conversation information E3 is set in the “target stack”. In the determination in step S104, since this candidate does not satisfy the attention condition “acquire municipalities”, the process proceeds to step S105. Also in step S105, since "Kagawa prefecture" is information of only prefectures, it is determined that all the conditions are not satisfied, so the process proceeds to step S106. In step S106, since there is no nested task in the operation corresponding to the condition "acquire municipalities", there is no task that satisfies the nested task. Therefore, the process proceeds to step S109. In step S1009, the candidate “Kagawa Prefecture” and the intention information “Kanagawa Prefecture” of the target stack are determined to be the same type of information, that is, the same address designation intention, so the notification destination corresponds to the dialog information E3. Task T6 to be performed. At this time, 30 points are subtracted from the score obtained by the natural language interpreting apparatus 1, and the score of “Kagawa” becomes 870.
[0163]
Subsequently, for the candidate “Kawasaki City”, the dialog information E3 is set in the “target stack” in step S101. In the determination in step S104, since “Kawasaki City” satisfies the attention condition “acquire municipalities”, the notification destination is determined to be the task T6 corresponding to the conversation information E3. At this time, the score of "Kawasaki City" remains at 880 points without deduction from the score obtained by the natural language interpreting apparatus 1.
[0164]
As a result of recalculating the score in consideration of the notification destination to the stack in this manner, the candidate “Kawasaki City” is adopted as the speech information as the user input. This can be said to be a selection that takes into account the situation where the system is inquiring about "city name". As described above, since the candidate selection unit 101 edits the score in consideration of the dialog stack state at that time, dynamic candidate selection according to the dialog state at that time becomes possible.
[0165]
Thus, according to the dialog management apparatus 100 configured as described above, even when it is determined that the topic is changed, the topic is interrupted without previously determining the return destination topic, and the topic contributes to the original topic. Even in such a case, it is possible to appropriately return to the original topic. Further, according to the present invention, it is possible to execute a candidate selection in consideration of a dialogue situation at the time of a user's input.
[0166]
Note that the present invention is not limited to the above embodiments. Although the plan managed by the dialog stack management unit 103 exemplifies that the tasks can be nested, the nested tasks may be described as being pushed as new dialog information. In this case, steps S106 and S107 of the determination operation in FIG. 3 will also be omitted.
[0167]
Note that, in the above-described example, the description is made as if implemented in the form of a car navigation system. However, these dialogue management methods are not limited to the above-described embodiment. The embodiment of the present invention is not limited to this. For example, the embodiment can be implemented by a tourist guide system or the like, and such an embodiment is also within the scope of the present invention. As described above, the embodiments of the present invention can be variously modified from the above-described examples, and they are also within the scope of the embodiments of the present invention unless they are contrary to the gist.
[0168]
As described above, according to the embodiment, when a task to be solved by the statement information obtained in the dialogue with the user is deleted from the stack, and task output information obtained as a result of solving the task is obtained, the task output information is stored in the stack. After the first task is stored in the stack, the task related to the task output information (herein referred to as the first task) is used (set) for the next conversation purpose. The assigned task is deleted from the stack, and the first task is set as the stack top. That is, when one problem is solved by the dialogue, as a result of the solution of the problem, the unnecessary problem which is no longer required as the purpose of the dialogue is searched based on the task output information of the solved problem, and the stack is searched. Remove from. Therefore, after one problem is solved, there is no longer a need to interact with a similar problem or a similar problem for the purpose of conversation, so that redundant and unnatural conversation can be avoided.
[0169]
Also, at least one condition for determining that one of the tasks stored in the stack (herein referred to as a first task) has been solved by the statement information obtained in the dialogue with the user. Then, it is determined that the first task is a task related to the utterance information, the task stored in the stack after the first task is deleted from the stack, and the first task is set to the conversation purpose. By setting to, when the purpose of the dialogue is changed by the user's intention, the purpose of the dialogue can be changed without a sense of incongruity. Can be avoided.
[0170]
(Second embodiment)
In the first embodiment, after the task is achieved, the destination of the output of the task (task output information) is searched for and the dialogue is advanced. However, the present invention is not limited to this case, and a new task is called. At this time, the notification destination of the task output information of the task may be predicted in advance. Therefore, as a second embodiment, when dialog information corresponding to a new task is stored in the stack, the dialog information is recorded in the stack based on task output information associated with the task. An interactive device that searches (predicts) a notification destination of task output information of a new task when a new task has been achieved from among tasks will be described.
[0171]
The configuration of the dialogue device according to the second embodiment is the same as the configuration of the dialogue device according to the first embodiment (see FIG. 1), and the operation of the dialogue stack operation unit 102 is different. In addition, when new conversation information is stored in the stack, the notification destination of task output information output when the problem corresponding to the new conversation information is solved is predicted in advance as described above. In order to record a notification destination as a prediction result, a notification possibility field is added to the conversation information.
[0172]
The above two points are different from those of the first embodiment, and the different points will be mainly described below.
[0173]
FIG. 26 shows an example of the state of the stack according to the second embodiment, which corresponds to the state of the stack shown in FIG. 13 of the first embodiment.
[0174]
As shown in FIG. 26, each dialogue information corresponds to the task used for the dialogue by the dialogue stack operation unit 102, as described above, and the content of the dialogue at a certain point in time is the stack top at that time (the top of the stack). (The above element).
[0175]
A "notification possibility" field is newly added to each piece of dialog information in FIG.
[0176]
The “notifiability” field of each piece of dialog information indicates a task that outputs the task output information when it is estimated that there is a possibility of a notification destination of the output of the dialog information (task output information) at the upper level. The fields to enumerate.
[0177]
In the stack shown in FIG. 26, according to the description of the “notifiability” field of the conversation information E2, the “location search” task T4 of the conversation information E2 includes the “location search by address” task corresponding to the conversation information E3. It is understood that it is estimated that the task output information of T6 is notified. Further, according to the description of the “notifiability” field of the dialog information E1, the “destination and place setting” task T2 of the dialog information E1 includes the task output information of the “location search” task T4 corresponding to the dialog information E2. It is understood that it is estimated that is notified. On the other hand, in the dialogue information E3, since there is no dialogue information in the upper stage, the “notifiability” field is empty. When multiple pieces of conversation information exist in the upper row and they are notified to the same conversation information, the plurality of pieces of conversation information are listed in the “possibility of notification” field.
[0178]
When loading a task, the dialog stack operation unit 102 adds the name of the task information loaded this time to the “notifiability” field of the dialog information of the load source task. When a new task is to be loaded from the utterance information obtained from the user's utterance content, the notification destination search process shown in FIGS. 4 and 5 is executed using the task output information of the outermost task to be loaded. As a result, if there is (a dialogue information corresponding to) the notification destination task, the task information to be newly loaded this time is notified to the “notifiability” field of the dialogue information. As described above, the load is a procedure that is called when proceeding to a more detailed topic in a dialog or when starting another topic.
[0179]
The transition is also a procedure for calling a new task, as described above. However, a new task is loaded so as to change the task used for the current conversation purpose. When the task name of the transition source is written in the “notifiability” field of the lower dialog information, the name of the task before the transition is added to the “notifiability” field of the dialog information. If the current task has been notified, the current task is notified to the conversation information. This is a procedure that is called when the conversation goes on in the order and the topic goes on next.
[0180]
FIG. 25 is a flowchart for explaining the entire processing operation of the dialog stack operation unit 102 according to the second embodiment. In FIG. 25, the same portions as those in FIG. 3 are denoted by the same reference numerals, and only different portions will be described. That is, in FIG. 3, when the task of a certain task is achieved and the task output information of the task is actually output, the task of the notification destination is searched for. Instead, in FIG. When the conversation information corresponding to a certain task is stored in the stack (when a new task is called by the utterance information, it is determined in step S14 of FIG. 25; In S10), a notification destination of task output information output from the task when the task of the task is achieved is estimated in advance. When the task of the task is achieved, in step S301, it is searched for a task estimated in advance as a notification destination of the task output information of the task by referring to the “notifiability” field of each piece of dialog information. For example, the task output information is notified to the task estimated as the notification destination.
[0181]
In step S2, a search is made for a notification destination of the utterance information stored in the buffer BF1. This notification destination search processing is the same as in FIGS.
[0182]
In step S301, it is checked whether there is a notification destination of the utterance information or the task output information stored in the buffer BF1. If there is a notification destination obtained in step S2 or the name of the task at the output source of the task output information is described below the conversation information corresponding to the task at the output source of the task output information, Search for dialogue information with a "possible" field. If there is a notification destination, the process proceeds to step S4; otherwise, the process proceeds to step S13.
[0183]
Next, as a specific example of the second embodiment, portions different from the specific example of the first embodiment will be described.
[0184]
(1) First remark: When “set a destination” is input to the natural language interpreting apparatus 1, the conversation stack operation unit 102 is notified of the remark information “destination setting”. The task corresponding to the utterance information “destination setting”, that is, the task T1 having the name of the utterance information “destination setting” is loaded, and the dialog stack operation unit 102 generates the dialog information E1 corresponding to the task T1. Push on the stack. At this time, since there is no dialog information stored below the dialog information E1 (dialog information stored in the stack before the dialog information E1 is stored in the stack), the notification of the task output information is performed when the task T1 is achieved. No previous search is performed. FIG. 27 shows the state of the stack at this time.
[0185]
In the same manner as described above, the task T4 is called (loaded) from the task T2, and the conversation information E2 corresponding to the task T4 is stored in the stack (Step S10). At this time, since the task T4 is called from the task T2 corresponding to the conversation information E1, the task output information of the task T4 is notified to the task T2 corresponding to the conversation information E1. Therefore, the “location search” task of the task T4 is added to the “notifiability” field of the interaction information E1. Since the task T4 is notified of "location search" stored in the buffer BF1, "location search" is written as the intention information of the dialog information E2 (step S4). FIG. 28 shows the state of the stack at this stage. Thereafter, in the same manner as described above, the target condition is selected, and the operation corresponding to the target condition is executed. FIG. 29 shows the state of the stack at this time.
[0186]
(2) When the stack is in the state shown in FIG. 29, it is assumed that a second statement: “search by address” is input to the natural language interpreting apparatus 1. It is assumed that the interpretation result of the natural language interpretation device 1 is “location search by address”. When this is notified to the dialogue stack operation unit 102, the statement information satisfies the attention condition of the dialogue information E2 in FIG. 29, so that the operation at the time of achieving all the conditions of the task T5 which is the current dialogue purpose is executed. (Step S6). As described above, the conversation information corresponding to the task T5 is deleted from the conversation information E2.
[0187]
In the operation when all the conditions are achieved, the "location search by address" task T6 is loaded, and as shown in FIG. 30, the conversation information E3 corresponding to the task T6 is pushed onto the stack (the conversation information E3 is the conversation information E3). Stacked on the information E2). At this time, since the task T6 has been loaded from the conversation information E2, the “location search by address” task T6 is added to the “notifiability” field of the conversation information E2.
[0188]
(3) Thereafter, in the same manner as described above, it is assumed that the third input: “Kanagawa Prefecture” is input and the stack is in a state as shown in FIG.
[0189]
(4) When the stack is in the state of FIG. 31, it is assumed that a fourth statement: “After all, with a registered name” is input to the natural language interpreting apparatus 1. Then, as shown in FIG. 32, the conversation information E4 is stored in the stack, and the notification destination of the task output information “searched place” of the “location search by registered name” task T7 corresponding to the conversation information E4 is set. In order to search, a notification destination search process shown in FIGS. 4 and 5 is performed. As a result, the task output information satisfies the condition of the “location search” task T4. Therefore, the “location search by registered name” task T7 is added to the “notifiability” field of the conversation information E2 corresponding to the task T4. (See dialog information E2 in FIG. 32).
[0190]
(5) Fifth remark: It is assumed that "Minato Mirai" is input to the natural language interpreting apparatus 1. When the comment information “location search by registered name: Minato Mirai” corresponding to this comment is input to the interactive device 100, the task of the task T7 corresponding to the interactive information E4 is solved by the comment information as described above (step). S5). The operation at the time of achieving all the conditions of the task T7 is executed, and “minatomirai” is output as the task output information. Then, the process proceeds to step S301 to search for the notification destination of the task output information with reference to the “notifiability” field of the conversation information in the stack, and the “location search by registered name” task T7 is designated. Information E2 is found. Therefore, the task output information stored in the buffer BF1 is notified of "minatomirai" to the "location search" task T4 corresponding to the dialog information E2.
[0191]
When the task output information “minatomirai” is notified to the “location search” task T4 corresponding to the conversation information E2, the conversation information E3 is popped from the stack by clearing. Further, since the condition of the “location search” task T4 is satisfied, the operation when all the conditions are achieved is executed, and “minatomirai” is output as the task output information. When this notification destination is searched in step S301, the "location search" task T2 has been added to the "notifiability" field of the dialog information E1, so that the task output information is notified to the task T2 of the dialog information E1. Is done. Then, the task of the task T2 is solved, and the operation at the time of achieving all the conditions of the task T2 is executed. Further, the conversation information corresponding to the task T2 whose problem has been solved in this way is deleted from the stack in the same manner as described above. FIG. 33 shows the state of the stack at this time.
[0192]
Since the transition to the task T3 is predetermined as the operation when all the conditions are achieved, the “destination approval” task T3 is loaded. The task output information of the task T3 is the searched and approved place. When the notification destination of the task output information is searched according to the notification destination search processing shown in FIGS. 4 and 5, the task output information satisfies the condition of the “destination setting” task T1 of the dialog information E1. A "destination approval" task T3 is added to the "notification possibility" field of the information E1. FIG. 34 shows the state of the stack at this time.
[0193]
(6) Sixth remark: When "yes" is input, the task of task T3 is solved in the same manner as described above. It notifies the task T1 estimated in advance as the information notification destination. Further, the dialog information corresponding to the task T3 is also deleted as described above. Since the task of the task T1 has also been achieved, the conversation information E1 corresponding to the task T1 is popped, but the stack becomes empty. Therefore, since there is no notification destination of the task output information of the task T1, the input waits here. FIG. 35 shows the state of the stack at this time.
[0194]
(7)) In the state where the stack is as shown in FIG. 34, when "Do you want to set this as the destination?" Is output, the sixth remark: the user changes the location instead of "Yes" It is assumed that a seventh statement: “At Yamashita Park” is intended to be performed, and this is input to the natural language interpreting apparatus 1.
[0195]
It is assumed that the interpretation result of the natural language interpretation device 2 is “location search by registered name: Yamashita Park” by referring to, for example, the conditions of each task of the plan stored in the plan storage unit 104. When this is notified to the dialog stack operation unit 102 as utterance information, the notification destination is searched in step S2, but all the conditions for solving the problem or the utterance information "location search by registered name: Yamashita Park" There is no dialog information of the same type as task and intention information that satisfies at least one of the plurality of conditions (steps S104 to S106 and step S109). Therefore, there is no task as a notification destination in the stack. In this case, since the new flag is “1” (step S13), the task corresponding to the utterance information, that is, the task T7 having the name of the information “location search by registered name” included in the utterance information is stored in the plan storage unit. Select from 104 and load. At this time, the conversation stack operation unit 102 generates the conversation information E5 corresponding to the task T7, and the conversation information E5 is pushed (the conversation information E5 is stacked on the conversation information E1).
[0196]
At this time, the notification destination of the task output information of the task T7 is searched. 4 and 5, “destination location setting” of task T2 included in task T1 corresponding to interaction information E1 is obtained as a notification destination. Therefore, “location search by registered name” of task T7 is added to the “notifiability” field of the dialog information. It is the same as described above that the utterance information is written in the intention information of the conversation information E5. FIG. 36 shows the state of the stack at this time.
[0197]
Now, since the statement information written as the intention information in the dialogue information E5 satisfies all the conditions for solving the task of the task T7, that is, since the registered name has been acquired (step S5), it is determined in advance in the task T7. The operation at the time of achieving all the conditions is executed (step S6). When the operation at the time of achieving all the conditions is executed, a location having a registered name of “Yamashita Park” (here, expressed as “yamashita-park”) is searched. Since the task output information predetermined for the task T7 is “location of search result”, “yamashita-park” which is a search result of the operation when all the conditions are achieved is output as task output information. The notification destination of this task output information is the conversation information E1 from the description in the “possibility of notification” field. More specifically, the task T2 is included in the task T1.
[0198]
That is, since the task T2 exists as the notification destination, the name of the dialogue information E1 is "destination setting-destination location setting", and the task output information stored in the buffer BF1 corresponds to the task corresponding to the dialogue information E1. Since T2 is notified, the intention information of the dialog information E1 is rewritten to "yamashita-park" instead of "minatomirai". When the task output information is notified to the task T2 corresponding to the dialog information E1 stored in the buffer BF1, all the conditions of the task T2 are satisfied (step S5). Therefore, information about the task T2 included in the task T1 is deleted from the conversation information E1. That is, in this case, "destination location" is deleted from the name of the conversation information E1. FIG. 37 shows the state of the stack at this time. At this stage, “yamashita-park” was obtained as the location of the search result.
[0199]
As described above, according to the second embodiment, when the conversation information corresponding to the new task (task) is stored in the stack, the notification destination of the task output information of the new task is set as the conversation information in the stack. Search from stored tasks. Then, when the new problem is actually solved, the task output information is notified to the previously searched task, and the task is used for the purpose of dialogue, thereby achieving the same effect as in the first embodiment. Play. The difference from the first embodiment is when to search for a notification destination of task output information. In the case of the first embodiment, the task output information is actually output. In the case of the form, it is when the task is called and the interaction information is stored on the stack.
[0200]
In the second embodiment, the task that outputs the received task output information is written in the notified destination field on the task output information receiving side. However, the present invention is not limited to this. For example, a “notification destination field” may be provided for each piece of conversation information to manage the notification destination from the notifying side, or to store tasks on both the notified side and the notifying side. Any method may be used as long as the notification destination is dynamically determined in consideration of the output.
[0201]
As described above, according to the first and second embodiments, the return destination when a certain task is interrupted and shifted to a new task is set as a result of the dialogue using the new task as a dialogue purpose. In response (by returning to the task associated with the task output information of the solved task), a flexible decision can be made.
[0202]
Even if the dialog using one task as the purpose of the conversation is interrupted (without solving the task) and changed to another task by the user's intention, after the other task is solved, Since a conversation related to another task or a similar task is not performed for the purpose of the conversation, redundant and unnatural conversation can be avoided.
[0203]
Even if it is determined that the topic is changed, the topic can be interrupted without determining the return destination topic in advance, and even if the topic contributes to the original topic, it can be properly restored to the original topic it can. In addition, it is possible to execute a candidate selection in consideration of a dialogue situation at the time of a user's input.
[0204]
The method of the present invention described in the embodiment of the present invention can be implemented as a program that can be executed by a computer, such as recording a magnetic disk (floppy disk, hard disk, etc.), an optical disk (CD-ROM, DVD, etc.), a semiconductor memory, or the like. It can also be stored in a medium and distributed.
[0205]
Note that the present invention is not limited to the above-described embodiment, and can be variously modified in an implementation stage without departing from the gist of the invention. Furthermore, the above embodiments include inventions at various stages, and various inventions can be extracted by appropriate combinations of a plurality of disclosed configuration requirements. For example, even if some components are deleted from all the components shown in the embodiments, at least one of the problems described in the column of the problem to be solved by the invention can be solved, and the effect of the invention can be solved. In the case where the effect described in (1) is obtained (at least one), a configuration from which this component is deleted can be extracted as an invention.
[0206]
【The invention's effect】
As described above, according to the present invention, when a dialogue is performed according to a plurality of predetermined tasks for achieving a predetermined object, when a certain task is interrupted and a new (another) task is transferred, The next task can be flexibly determined according to the result of the dialogue with the new task.
[Brief description of the drawings]
FIG. 1 is an exemplary view showing a configuration example of a dialogue device according to an embodiment of the present invention.
FIG. 2 is a view for explaining plans stored in a plan storage unit.
FIG. 3 is a flowchart illustrating a processing operation of a dialog stack operation unit.
FIG. 4 is a flowchart for explaining a notification destination search processing operation among processing operations of a dialog stack operation unit.
FIG. 5 is a flowchart for explaining a notification destination search processing operation among processing operations of a dialog stack operation unit.
FIG. 6 is a diagram showing a state of a stack.
FIG. 7 is a diagram showing a state of a stack.
FIG. 8 is a diagram showing a state of a stack.
FIG. 9 is a diagram showing a state of a stack.
FIG. 10 is a diagram showing a state of a stack.
FIG. 11 is a diagram showing a state of a stack.
FIG. 12 is a diagram showing a state of a stack.
FIG. 13 is a diagram showing a state of a stack.
FIG. 14 is a diagram showing a state of a stack.
FIG. 15 is a diagram showing a state of a stack.
FIG. 16 is a diagram showing a state of a stack.
FIG. 17 is a diagram showing a state of a stack.
FIG. 18 is a diagram showing a state of a stack.
FIG. 19 is a diagram showing a state of a stack.
FIG. 20 is a diagram showing a state of a stack.
FIG. 21 is a diagram showing a state of a stack.
FIG. 22 is a diagram showing a state of a stack.
FIG. 23 is a diagram showing a state of a stack.
FIG. 24 is a diagram showing a state of a stack.
FIG. 25 is a flowchart for explaining another processing operation of the dialog stack operation unit.
FIG. 26 is a diagram for explaining information stored as conversation information.
FIG. 27 is a diagram showing a state of a stack.
FIG. 28 is a diagram showing a state of a stack.
FIG. 29 is a diagram showing a state of a stack.
FIG. 30 is a diagram showing a state of a stack.
FIG. 31 is a diagram showing a state of a stack.
FIG. 32 is a diagram showing a state of a stack.
FIG. 33 is a diagram showing a state of a stack.
FIG. 34 is a diagram showing a state of a stack.
FIG. 35 is a diagram showing a state of a stack.
FIG. 36 is a diagram showing a state of a stack.
FIG. 37 is a diagram showing a state of a stack.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Natural language interpretation apparatus, 2 ... Response output apparatus, 100 ... Dialogue apparatus, 101 ... Candidate selection part, 102 ... Dialogue stack operation part, 103 ... Dialogue information storage part, 104 ... Plan storage part

Claims (17)

An interaction method for interacting with a user to solve a plurality of problems for achieving a predetermined purpose,
A first step of, among the plurality of tasks, storing unresolved tasks selected for the purpose of interacting with the user in the storage unit in the order of selection;
A second step of obtaining the first information by using the unresolved task last stored in the storage means for the purpose of dialogue;
A third step of solving a first unresolved problem stored in the storage means with the first information and obtaining second information;
A fourth step of deleting the solved first unsolved problem from the storage means;
The dialogue after the second unresolved task is used for using the second unresolved task related to the second information stored in the storage means for the next conversational purpose in the second step. A fifth step of deleting the unresolved issues selected for the purpose from the storage means;
An interactive method, comprising: The second unresolved problem related to the second information is characterized in that the second information satisfies at least one condition for determining that the second unresolved problem has been solved. The method according to claim 1. When the second unresolved issue related to the second information is not stored in the storage unit, the first unresolved issue among the unresolved issues stored in the storage unit is used for the conversation purpose. 2. The dialogue method according to claim 1, wherein an unsolved task selected for the dialogue just before the selection is used for the next dialogue in the second step. 2. The interactive method according to claim 1, wherein the storage unit is a stack that stores unsolved tasks selected for the purpose of interacting with the user in the order of selection. When the first unresolved problem solved by the first information is not stored in the storage unit, a third unresolved problem related to the first information is stored in the second step. A sixth step of deleting, from the storage means, the unresolved issues selected for the dialogue purpose after the third unresolved task for use for the dialogue purpose of the above;
When a third unresolved issue related to the first information is not stored in the storage unit, a new issue related to the first information among the plurality of issues is set as the unresolved issue. A seventh step of storing in said storage means for use in the second step for the next said interactive purpose;
The interactive method according to claim 1, further comprising: The second step is
When a plurality of candidates for the first information estimated from the user's input are given, the order of use of the unresolved issues stored in the storage means for the purpose of the dialogue among the plurality of candidates The method according to claim 1, wherein the first information is selected as the first information by giving priority to a candidate relating to an unresolved task that comes after the priority. The method according to claim 1, wherein when the second information is obtained in the third step, the second unresolved task is obtained. When the first unresolved task selected for the dialogue purpose in the first step is stored in the storage unit, the second information obtained when the first unresolved task is solved is estimated. An eighth step of previously obtaining the second unresolved issue based on this,
The interactive method according to claim 1, wherein in the fifth step, the second unsolved task obtained in advance is used for the next interactive purpose in the second step. A dialogue device for performing a dialogue with a user to solve a plurality of problems for achieving a predetermined purpose,
Storage means for storing, among the plurality of tasks, unresolved tasks selected for the purpose of interacting with the user in the order of selection;
First means for obtaining first information by using the unresolved task lastly stored in the storage means for interactive purposes;
A second means for solving the first unresolved problem stored in the storage means by the first information and obtaining second information;
Third means for deleting the solved first unsolved problem from the storage means;
A dialogue after the second unresolved task, wherein the second unresolved task related to the second information stored in the storage means is used by the first means for the next conversational purpose. Fourth means for deleting the unresolved issues selected for the purpose from the storage means;
An interactive device, comprising: The second unresolved problem related to the second information is characterized in that the second information satisfies at least one condition for determining that the second unresolved problem has been solved. The interactive device according to claim 9. When the second unresolved issue related to the second information is not stored in the storage unit, the first unresolved issue among the unresolved issues stored in the storage unit is used for the conversation purpose. 10. The interactive apparatus according to claim 9, wherein an unsolved task selected for the interactive purpose immediately before the selection is used for the next interactive purpose by the first means. 10. The interactive apparatus according to claim 9, wherein the storage unit is a stack that stores unsolved tasks selected for the purpose of interacting with the user in the order of selection. When the first unresolved problem solved by the first information is not stored in the storage unit, a third unresolved problem related to the first information is stored in the first unit. Fifth means for deleting, from the storage means, the unresolved issues selected for the dialogue purpose after the third unresolved task for use for the dialogue purpose of
When a third unresolved issue related to the first information is not stored in the storage unit, a new issue related to the first information among the plurality of issues is set as the unresolved issue. Sixth means for storing in said storage means for use by said first means for the next said interaction purpose;
The interactive device according to claim 9, further comprising: The first means includes:
When a plurality of candidates for the first information estimated from the user's input are given, the order of use of the unresolved tasks stored in the storage unit used for the dialogue purpose among the plurality of candidates 10. The interactive apparatus according to claim 9, wherein a candidate related to an unresolved task that comes after is preferentially selected as the first information. 10. The interactive device according to claim 9, wherein when the second information is obtained by the second means, the second unsolved task is obtained. When the first unresolved task selected for the dialogue purpose is stored in the storage unit, the second information obtained when the first unresolved task is solved is estimated, and based on this, Further comprising a seventh means for determining the second unresolved problem,
10. The interactive apparatus according to claim 9, wherein the fourth means uses the second unresolved task obtained in advance for the next interactive purpose by the first means. A program for interacting with a user to solve a plurality of problems for achieving a predetermined purpose,
On the computer,
A first step of, among the plurality of tasks, storing unresolved tasks selected for the purpose of interacting with the user in the storage unit in the order of selection;
A second step of obtaining the first information by using the unresolved task last stored in the storage means for the purpose of dialogue;
A third step of solving a first unresolved problem stored in the storage means with the first information and obtaining second information;
A fourth step of deleting the solved first unsolved problem from the storage means;
The dialogue after the second unresolved task is used for using the second unresolved task related to the second information stored in the storage means for the next conversational purpose in the second step. A fifth step of deleting the unresolved issues selected for the purpose from the storage means;
A program that executes

Images