JP7117951B2 - Dialogue plan creation support device, dialogue plan creation support method, and program - Google Patents

Description

The present invention relates to a dialog plan creation support device, a dialog plan creation support method, and a program.

Conventionally, an input-response interactive system is known that outputs a system response for each user's input. Typical examples include question answering systems and recent AI speakers.

Also known is a scenario-based dialogue system that conducts dialogue according to a predetermined scenario in order to achieve the purpose of a certain task. Examples include a dialog system for airplane ticket reservations and robots placed at stores to elicit requirements from visitors.

Japanese Patent Laid-Open No. 2002-200001 discloses a voice dialogue system that, in voice dialogue, responds by understanding the meaning of even if the user's utterance is a short word. The voice dialogue system described in Patent Document 1 prepares a dialogue scenario in which three contents of a first system utterance, a user utterance, and a second system utterance are combined. , is an invention that returns a natural response considering the content of the previous system utterance regardless of the length of the user's utterance.

JP 2018-54790 A

In the voice dialogue system as described above, the more dialogue scenarios there are, the more natural dialogue can be realized. SUMMARY OF THE INVENTION In view of the above background, it is an object of the present invention to provide a dialogue plan creation support device that assists a plurality of workers in simultaneously creating a dialogue scenario in a distributed parallel manner.

A dialogue plan creation support device according to the present invention is a dialogue plan composed of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user. A device for supporting the creation of a dialog plan including information on the transition destination of the communication that communicates with a dialog plan database that stores the dialog plan and a worker terminal that a worker creates and inputs the dialog plan. a dialogue plan creation support screen is transmitted to the worker terminal via the communication unit, the data of the dialogue plan input on the creation support screen is received, and the received data of the dialogue plan is transmitted to the interaction plan a controller for storing in a plan database.

In this way, the data of the dialogue plan input to the worker terminal is received via the communication unit and stored in the dialogue plan database, so that a large number of dialogue plans can be created by a plurality of workers. In recent years, a form of outsourcing work to an unspecified number of human resources via the Internet, so-called crowdsourcing, has emerged. According to the configuration of the present invention, creation of dialogue plans can be entrusted to many crowd workers, and a wide variety of dialogue plans can be easily generated.

In the dialogue plan creation support device of the present invention, when the data of the dialogue plan received from the worker terminal is stored, the control unit gives the dialogue plan an undetermined status, , the undetermined status may be changed to determined. In this way, the quality of the dialogue plan can be maintained by having the manager check the dialogue plan instead of adopting the dialogue plan created by the worker as it is. Also, by managing the dialogue plans with state status, it is possible to use the dialogue plan with the confirmed status in the production system and use the dialogue plan with the undetermined status at the time of testing. can.

In the dialog plan creation support device of the present invention, the creation support screen may display candidates for the dialog plan to be transitioned to. With this configuration, it is possible to easily select the dialogue plan of the transition destination, and to support the creation of the dialogue plan.

In the dialogue plan creation support device of the present invention, candidates for transition destination dialogue plans may be grouped on the creation support screen. For example, group dialogue plans on the same topic. With this configuration, by selecting a group of dialogue plans, it is not necessary to select a plurality of dialogue plans one by one, which is efficient.

In the dialogue plan creation support device of the present invention, the creation support screen may display the number of backlinks to which dialogue plan candidates of transition destinations are linked as transition destinations. By displaying the number of linked dialogue plans in this way, it is possible to prevent too many transitions to a particular dialogue plan.

In the dialogue plan creation support device of the present invention, the creation support screen for creating the user response has an input item for a regular expression pattern of a response from the user and an input item for an example response of the regular utterance pattern. Then, when an update button provided on the creation support screen is selected, the control unit may determine whether the response example matches the regular expression pattern. Input of a regular expression pattern must comply with predetermined grammatical rules, but according to the configuration of the present invention, it is possible to easily determine whether or not the regular expression pattern is correct by inputting a response example.

In the dialogue plan creation support device of the present invention, the creation support screen has a button for saving the user response created by the user as a fixed expression, and when the save button is selected, the control unit The expression may be stored in a storage unit, and it may be described that the fixed expression stored in the storage unit should be referred to. With this configuration, it is possible to easily generate a user response by referring to the fixed expression stored in the storage unit, and to correct the fixed expression stored in the storage unit so that the user response of the dialogue plan referring to this can be easily generated. can be modified in batches.

In the dialogue plan creation support device of the present invention, the creation support screen includes check boxes for selecting the created user response and the received utterance for that user response, and the user response selected in another plan and and a button to copy the received utterance. With this configuration, the user response and the corresponding received utterance can be used.

In the dialogue plan creation support device of the present invention, the creation support screen may have an item for selecting whether or not the created dialogue plan can be used as an initial utterance. By having an item for selecting whether or not to be used as an initial utterance in this way, a natural initial utterance can be realized by excluding an unnatural dialogue plan from the initial utterance when uttered first.

In the dialogue plan creation support device of the present invention, the creation support screen may have an item for selecting whether or not the created dialogue plan can be used more than once in a series of dialogues. By having an item for selecting whether or not the dialogue plan can be used more than once in a series of dialogues in this way, it is possible to set this input item without forgetting. A natural dialogue can be realized by specifying an unnatural dialogue plan that is uttered multiple times.

In the dialog plan creation support device of the present invention, the creation support screen may have an item for selecting whether or not to end the dialog without designating the transition destination of the created dialog plan. This configuration allows the creation of an interaction plan that terminates the interaction.

In the dialogue plan creation support device of the present invention, the control unit transmits the creation support screen including a test button for executing a test of the interaction plan being created, and when the test button is operated, the interaction plan test may be performed. With this configuration, it is possible to confirm whether or not the interaction plan being created functions without discomfort.

The dialogue plan creation support device of the present invention may cause the operator to input a user response to the pretended utterance in the test. This configuration allows testing of possible responses for correct transitions. The operator may select a user response from example responses or enter a new user response. This makes it possible to test patterns that have user responses that are not registered in the interaction plan.

In the dialogue plan creation support device of the present invention, the control unit may hierarchically display the dialogue plan on the display unit according to the transition destination information. With this configuration, the flow of the dialogue plan can be viewed at a glance, and the hierarchy can be confirmed.

In the dialogue plan creation support device of the present invention, the control unit may output a warning when there is a looping portion in the transition of the dialogue plan. When multiple workers set links in the dialogue plan, unintended loops may occur. According to the configuration of the present invention, the loop can be eliminated by notifying that there is a loop by outputting a warning.

In the dialogue plan creation support device of the present invention, the control unit may output the number of times to return to the same dialogue plan. This can help determine whether loops are allowed.

In the dialogue plan creation support device of the present invention, the control unit uses the dialogue plan stored in the dialogue plan database, starting from one selected dialogue plan, to reach the closing utterance or to the transition destination. The dialogue may be simulated until there is no more, the number of continued dialogue plans may be counted, and the result may be displayed on the display unit. With this configuration, it is possible to visualize the number of exchanges that continue without the dialogue breaking down, and to evaluate the quality of the dialogue plan. In addition, based on this result, for example, it becomes easier to establish a policy that utterance content or response content should be supplemented for topics on which dialogue does not continue.

In the dialogue plan creation support device of the present invention, the control unit may count the number of times each dialogue plan is used, and display the result on the display unit. As a result, it is possible to grasp frequently used dialogue plans and less frequently used dialogue plans, and to review the balance of the whole dialogue plan.

In the dialogue plan creation support device of the present invention, the control unit may cause the display unit to display, as reasons for ending the dialogue, the number of times a closing utterance was reached and the number of times there was no transition destination. With this, it is possible to grasp whether the dialogue ended normally or not.

In the dialogue plan creation support device of the present invention, the dialogue plan includes a general user response that can be used regardless of the content of the utterance, and the control unit counts the number of matches with the general user response. It may be displayed on the display unit. If it matches a general user response, it is considered that an appropriate user response has not been set, so the configuration of the present invention makes it possible to check the validity of the dialogue plan.

A dialogue plan creation support method according to the present invention is a dialogue plan composed of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user. A method for supporting the creation of a dialog plan including information on the transition destination of the, wherein a step of transmitting a dialog plan creation support screen to a worker terminal where a worker creates and inputs the dialog plan; a step of accepting input of dialogue plan data through the creation support screen at the worker terminal; receiving the dialogue plan data inputted at the worker terminal; storing the received dialogue plan data in a dialogue plan database; and storing.

The program of the present invention is a dialogue plan composed of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user. A program for supporting the creation of an interaction plan containing information, the step of sending to a computer an interaction plan creation support screen to a worker terminal where the worker creates and inputs the interaction plan. a step of receiving input of dialogue plan data through the creation support screen at the worker terminal; receiving the dialogue plan data inputted at the worker terminal; to execute the step of storing in the

According to the present invention, a plurality of workers can concurrently create a dialogue plan in a distributed parallel manner.

1 is a diagram showing the configuration of a dialogue plan creation support system according to an embodiment; FIG. It is a figure which shows the example of the data memorize|stored in interaction plan DB. It is a figure which shows an example of an interaction plan. It is a figure which shows an example of an interaction plan. It is a figure which shows an example of an interaction plan. It is a figure which shows an example of an interaction plan. It is a figure which shows the whole structure of a preparation support screen. FIG. 11 is a diagram showing an example screen of a dialogue plan list; FIG. 10 is a diagram showing an example of a dialog plan editing screen; It is a figure which shows the example of a screen of shaking utterance edit. FIG. 10 is a diagram showing an example screen for editing a user response; FIG. 10 is a diagram showing an example screen for editing a user response; FIG. 11 is a diagram showing an example of a received utterance editing screen; FIG. 10 is a diagram showing an example of a list for selecting a transition destination; FIG. (a) is a diagram showing an example of a list for selecting a transition destination group. (b) is a diagram showing an example of a screen for grouping dialogue plans. It is an example of the plan edit screen of the interaction plan of the copy source. It is an example of the plan edit screen of the interaction plan of the copy destination. It is an example of the plan edit screen of the interaction plan of the copy destination. FIG. 10 is a diagram showing an example of a dialogue plan test screen; FIG. 10 is a diagram showing an example of a dialogue plan test screen; FIG. 10 is a diagram showing an example of a dialogue plan test screen; FIG. 10 is a diagram showing an example of a dialogue plan test screen; FIG. 10 is a diagram showing an example of a screen displaying results of an automatic test of a dialogue plan; It is a figure which shows the example of a screen of a batch test. (a) It is a graph which shows the result of the closing utterance arrival number. (b) It is a graph which shows the result of the exception passing number. (c) is a graph showing the results of no-transition counts; It is a figure which shows the example which carried out hierarchical display of the plan list.

DESCRIPTION OF THE PREFERRED EMBODIMENTS A dialogue plan creation support device according to an embodiment of the present invention will be described below with reference to the drawings.
FIG. 1 is a diagram showing the configuration of a dialogue plan creation support system 1 according to an embodiment. The dialogue plan creation support system 1 has a dialogue plan creation support device 10 and a plurality of work terminals 20 connected to the dialogue plan creation support device 10 via the Internet. The dialogue plan creation support system 1 of the present embodiment assumes that a plurality of crowd workers (herein referred to as "workers" in this specification) are entrusted with the task of creating a dialogue plan in a business form called crowdsourcing. ing. The worker creates an interaction plan using the work terminal 20 and transmits it to the interaction plan creation support device 10 .

In order to appropriately continue the dialogue based on the dialogue plan, a huge dialogue plan is necessary. The dialogue plan creation support device 10 assists a plurality of workers to create a consistent dialogue plan. The dialogue plan creation support device 10 transmits a creation support screen to the work terminal 20 used by the worker, and assists the creation of the dialogue plan through the interface of the creation support screen. The creation support screen includes the screen itself and a script for controlling operations within the screen.

The dialogue plan creation support device 10 has an operation unit 11 , a control unit 12 , a communication unit 13 , and a dialogue plan database (hereinafter referred to as “dialogue plan DB”) 14 . The operation unit 11 has a function of receiving operations from an administrator. The control unit 12 transmits the creation support screen data to the work terminal 20 through the communication unit 13, inputs the data of the dialogue plan using the creation support screen, and transmits the data of the dialogue plan transmitted from the work terminal 20 through the communication unit 13 as the dialogue plan. Store in DB14.

FIG. 2 is a diagram showing an example of data stored in the dialogue DB 14. As shown in FIG. The dialogue plan DB 14 stores a plurality of dialogue plans. Each interaction plan has a dictated utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user. An ID for identifying the dialogue plan (hereinafter referred to as “dialogue plan ID”) is assigned to the dialogue plan. The dialogue plan ID is automatically numbered by the dialogue plan creation support device 10 .

The dialogue plan has a transition destination ID that defines which dialogue plan to transition to after the received utterance. The transition destination ID is the dialogue plan ID of the transition destination. Thereby, the dialogue system can easily determine which dialogue plan should be preferentially transitioned to after executing one dialogue plan ID. As the transition destination ID, a plurality of transition destination dialogue plan IDs are usually defined. Which dialogue plan is selected as a transition destination from a plurality of dialogue plans can be determined based on the situation of dialogue, user's preference, and the like.

Since the transition destination ID is information defining a dialogue plan to which transition should be preferentially made, the dialogue system can transition to a dialogue plan not defined in the transition destination ID. For example, if the score obtained when the dialogue plan specified by the transition destination ID is used as the transition destination does not reach a predetermined threshold value, the transition destination dialogue plan may be determined from among all the dialogue plans.

Also, the dialogue plan stored in the dialogue plan DB 14 has a status indicating whether the dialogue plan is finalized or undetermined. The state in which the manager has not yet checked the interaction plan created by the worker is the unconfirmed status, and the state in which the manager has confirmed the contents is the confirmed status. The administrator, by operating the operation unit 11, reads out the unconfirmed dialogue plan from the dialogue plan DB 14, confirms the contents, and then changes the unconfirmed status to the confirmed status. Here, an example of having two statuses, confirmed and unconfirmed, is given, but more statuses may be added, such as, for example, waiting for compatibility test, compatibility test completed, final confirmed, and the like.

An example of the hardware of the dialogue plan creation support device described above is a computer equipped with a CPU, RAM, ROM, hard disk, display, keyboard, mouse, communication interface, and the like. By storing a program having modules for realizing the functions described above in the RAM or ROM and executing the program by the CPU, the above-described dialogue plan creation support device is realized. Such programs are also included in the scope of the present invention. Next, a specific example of the dialogue plan will be described.

(Specific example of dialogue plan)
3 to 6 are diagrams showing examples of dialogue plans. FIG. 3 is a diagram showing an overview of the dialogue plan, and describes the utterances, user responses, and received utterances included in the dialogue plan. The dialogue plan with "plan ID: P101" shown in FIG. 3 has a domain of "gourmet", a subdomain of "meal", and a plan name of "Sake_1". The outline of the plan is ``pretend to dig deep into the contents of dishes that go well with alcoholic beverages''.

In this dialogue plan, the possibility of initial speech is set to "impossible", and it cannot be used when starting a dialogue with the user. This dialogue plan is a dialogue plan that is selected when a keyword such as a dish name appears after dialogue with the user has already been performed. Pretended utterances for which initial utterances are possible include, for example, "Where are you going?" and "What are you going to eat tonight?" be.

In addition, the reusability of this dialogue plan is "impossible", and this dialogue plan cannot be used twice in a series of dialogues with the user. Reusable pretend utterances include "What other fruits do you like?" and "Which other countries have you been to?" Since there is more than one response to such a pretended utterance, there is no problem in asking multiple times.

The outline of the pretend utterance is "pretend to delve into dishes that go well with alcohol", and examples of utterances are "Gyu-suji wants alcohol, right?" is. What is shown in FIG. 3 is an example, and in reality, the portions of "beef muscle" and "sake" are replaced with words that appear in dialogue with the user. Details will be described with reference to FIG.

"User Response→Received Utterance" indicates an outline of the user's response to the shaking utterance and an example of its utterance, and an outline of the received utterance to the user's response and an example of its utterance. Again, what is shown in FIG. 3 is an example and, in practice, contextual utterances are made.

FIG. 4 is a diagram showing the contents of the swing utterance of the dialogue plan of "plan ID: P101". The outline of the plan is, as shown in FIG. 3, "pretend to delve into the contents of dishes that go well with alcohol", and the dialogue act is "question".

The utterance format of the swinging utterance is "<s1> makes you want <s2>, isn't it?" It makes me want to drink.” The utterance example is a specific example because it may be difficult to understand the contents of the utterance format alone. In other words, the example utterances are not essential for processing in the dialogue system 1, but having the example utterances facilitates maintenance and the like.
In the utterance format, <s1><s2> is a query index, and contains values obtained through interaction with the user and values related thereto.

"Condition" is a condition for selecting this swinging utterance. The conditions include the conditions for the dialogue act and the conditions for the contents of the dialogue. As for the dialogue act, it is a condition that the immediately preceding dialogue act "includes" an "opinion/impression". Here, the condition is that "opinions/impressions" are "included", but various other variations are conceivable. It is also possible to make it a condition that "includes". In addition, when the condition is "does not include", it is also possible to set the condition not to include not only "question" but also a plurality of dialogue forms.

In this example, the conditions for the contents of the dialogue are defined as "<dish name> present in the dialogue base data" and "<sake> having a good compatibility with <s1> obtained from the cooking knowledge database". ing. The dialogue-based data will be described later, but in simple terms, it is the data of the concept that appeared in the dialogue with the user. The dialogue-based data is stored and managed in the dialogue-based data storage unit 22 .

The condition "<dish name> exists in the dialogue-based data" is that <dish name> is included in the dialogue with the user so far by referring to the dialogue-based data, If it is included, this query sentence is determined to be "true" and <dish name> is substituted for <s1>. If <dish name> does not exist in the dialogue-based data, the result of this query sentence will be "false" and this pretend utterance will not be applied. The number or ratio of query sentences determined to be “true” may be represented by a score, and a pretended utterance with a high score may be selected. If there are multiple dish names in the dialogue-based data, the utterance that satisfies the subsequent conditions and has the highest score is selected.

The condition "<sake> that has a good compatibility with <s1> obtained from the cooking knowledge service" is based on the information obtained from the cooking knowledge service. It's a good thing. The cooking knowledge service has a score that indicates the compatibility of food, and the compatibility is determined based on this score. If there is good compatibility, this query sentence is determined to be "true", and <sake> and its score are substituted for <s2>.

If the user eats a dish that goes well with alcohol, all of the above query sentences will be "true", and there is a high possibility that this pretended utterance will be selected. The number or ratio of query sentences determined to be “true” may be represented by a score, and a pretended utterance with a high score may be selected. In FIG. 4, the conditions regarding the data stored in the dialogue-based data have been described as an example, but data regarding user preferences, which will be described later, may be used as query conditions. As a result, it is possible to make an utterance that matches the user's taste.

When the swing utterance shown in FIG. 4 is uttered, the query index < Enter “dish name” in s1> and “sake” in <s2>. As a result, in response to the user's utterance that he or she ate the food, the user can say, "<Food name> makes you want alcohol, doesn't it?" Able to make utterances while preserving context. It should be noted that it is possible to select at random which one of "<dish name> makes me want alcohol" and "<dish name> makes me want to drink more".

FIG. 5 is a diagram showing one of candidate user responses for the dialogue plan with "plan ID: P101". The outline of the plan shown in FIG. 5 is the "positive response" and the dialogue act is the "answer". Although FIG. 5 shows an example of a positive response, "plan ID: P101" has, for example, "negative response" and "silent response" as user response candidates. The utterance regular expression pattern of the user response is "That's right | That's right | Yeah | Matches | Delicious | Yes | YES | is interpreted as a positive response.

"Condition" is a condition for selecting this swinging utterance. The conditions include the conditions for the dialogue act and the conditions for the contents of the dialogue. This content is the same as that explained in the swinging utterance shown in FIG.

As for "dialogue-based", "based data added as a compatible relationship between <s1> and <s2>" means that "<s1> makes you want <s2> in the pretended utterance. In response to the affirmative response from the user, it was agreed that <s1> and <s2> are a good combination, so <s1> and The fact that <s2> is in a compatible relationship is additionally recorded in the dialogue-based data. As a result, the dialogue can be appropriately continued using the dialogue-based data.

The "transition destination received utterance pattern" defines candidates for the transition destination of the received utterance from the system following the user response. Summaries include "praise/evaluation of hand-made" and "response with dish name in pretended utterance → response associated with restaurant", etc., and examples of received utterances are shown for each. As described above, the received utterance example is not essential for the processing of the dialogue system 1, but having the utterance example facilitates maintenance and the like. Which transition destination is to be selected from a plurality of transition destination received utterance pattern candidates is determined by matching the conditions of each candidate with the dialogue base data.

FIG. 6 is a diagram showing one candidate for the received utterance of the dialogue plan with "plan ID: P101". The outline of the received utterance shown in FIG. 6 is "Response associated with dish name in swing utterance→Response associated with alcohol", and the dialogue act is "Impression/opinion response".

In this example, there are three "utterance formats" for the received utterance: "That's going to be <s2>", "I'm wondering if it's Japan <s2> or red wine", and "It's perfect for <s2>". is attached with example utterances.

"Condition" is a condition for selecting this received utterance. The conditions include the conditions for the dialogue act and the conditions for the contents of the dialogue. As for the dialogue act, it is a condition that the immediately preceding dialogue act "includes" the "answer".

In this example, the conditions for the contents of the dialogue are "relationship that <s1> and <s2> existing in the underlying data are compatible" and "<s1> and <s2> existing in the underlying data". relationship” is stipulated. This received utterance is more likely to be selected when these conditions are "true".

"Transition destination dialogue plan" defines the transition destination plan ID of the dialogue plan to which transition should be preferentially performed after "Plan ID: P101". In this example, "gourmet_meal_stimulating food_1" is the preferential transition destination. The transition destination dialogue plan defined here is a preferential transition destination, and if the dialogue situation does not meet the conditions of the transition destination plan, the transition to the transition destination plan ID is not necessarily made. In that case, the dialogue control unit 1213 searches for a dialogue plan that meets the conditions from other dialogue plans, and determines it as the transition destination.

(Creation support screen)
Next, a creation support screen for assisting the creation of the dialogue plan as described above will be described.
FIG. 7 is a diagram showing the overall configuration of the creation support screen. The creation support screen has a dialogue plan list screen, dialogue plan edit screen, swing utterance edit screen, user response edit screen, and received utterance edit screen. The dialogue plan creation support device 10 supports the creation of the dialogue plan on the work terminal 20 by transmitting these screens to the work terminal 20 .

The dialogue plan list screen is a screen that displays a list of dialogue plans created so far. The dialogue plan edit screen is a screen for editing a plan selected on the dialogue plan list screen or a newly created dialogue plan. The shaking utterance edit screen, the user response editing screen, and the received utterance editing screen are screens for creating the shaking utterance, user response, and received utterance, respectively.

From the dialogue plan list screen, by selecting one of the displayed dialogue plans or by selecting new creation, the dialogue plan editing screen is displayed. From the dialog plan edit screen, it is possible to transition to each of a swing utterance edit screen, a user response edit screen, and a received utterance edit screen. It is possible to switch between the shaking utterance edit screen and the user response edit screen, and it is also possible to switch between the user response edit screen and the received utterance edit screen. The screen transition is performed by accessing the interaction plan creation support device 10 and downloading the corresponding creation support screen.

FIG. 8 is a diagram showing a screen example of a dialogue plan list. The dialogue plan list screen has a "narrow down" input section for narrowing down the dialogue plans to be displayed as a list from among a large number of dialogue plans. Narrowing down can be performed by each item of "domain", "plan name", "plan outline", and "person in charge".

The dialogue plan list screen displays a list of dialogue plans narrowed down by the narrowing-down conditions. The dialog plan list screen displays the "domain", "plan name", "plan overview", "status", and "person in charge" of the dialog plan. There is a display of "New" on the dialogue plan list screen, and a link to the dialogue plan edit screen is pasted on this display. Clicking on "Create new" opens the dialogue plan editing screen. At this time, default contents may be described on the interaction plan editing screen. For example, a template of an outline of the dialogue plan may be prepared, the user may select a template, and the content of the selected template may be displayed in the initial state of the dialogue plan edit screen. When a plan name is clicked in the dialogue plan list, the dialogue plan edit screen is opened with the contents of the clicked dialogue plan entered. By changing and copying the contents on the dialogue plan edit screen on which the contents have been input, it is possible to efficiently create similar dialogue plans.

FIG. 9 is a diagram showing an example of a dialog plan editing screen. FIG. 9 is an interactive plan whose plan name is "listen to favorite food_1".
At the top of the dialogue plan editing screen, there are buttons for "update", "copy", "delete" and "paste user response". Note that the "Paste User Response" button cannot be selected at this time. The "Update" button has a function of saving the latest content edited on the interaction plan editing screen. A "delete" button is a button for deleting this interaction plan. The buttons for "copy creation" and "user response paste" will be described later.

At the top of the interactive plan edit screen, "To plan list >> To plan test" is displayed, and links are pasted to each. If you select "To plan list", you will return to the plan list editing screen. When "to plan test" is selected, a transition is made to a test screen for testing this interaction plan. The test screen will be described later.

The "rule type" in the "plan details" column is data indicating the pattern of this dialogue plan. Here, "Y: Adjacent triple pattern" indicates that the dialogue plan consists of adjacent triples of the utterance, the user's response, and the received utterance. The dialogue plan edit screen also supports creation of dialogue plans for patterns other than the adjacent triple pattern. Examples of rule types include "N: adjacent pair pattern", "SQ: static question-answer pattern", "DQ: dynamic question-answer pattern", "A: connecting utterance pattern", "R: request pattern", and "U: arbitrary response". pattern” and “E: Closing utterance pattern”. Here, the connecting utterance pattern is an utterance pattern of a short fixed phrase to be sandwiched between the user response and the system utterance, the request pattern is an utterance pattern for issuing a command to the system, and the arbitrary response pattern is: Speech patterns emitted by the system regardless of user response. A closing utterance pattern is an utterance pattern at which the transition of the dialogue plan ends.

"Plan ID" is an ID that identifies the dialogue plan, and "Domain" is the domain to which the dialogue plan belongs. The "plan name" is the name of the dialogue plan, and the "plan overview" is a specific description of the contents of the dialogue plan.

“Initial utterance availability” is an item for selecting whether or not the dialog plan being edited can be used as an initial utterance. By setting this item to "possible", it becomes possible to use it as the first swinging utterance from the system to the user. A dialogue plan that allows an initial utterance is a utterance that can be started without prior knowledge, and is a case of asking "Do you have a favorite food?" like this dialogue plan. In the case of a dialogue plan that cannot be used as an initial utterance, the 'initial utterance availability' is set to 'no'.

“Reusability” is an item for selecting whether or not the dialogue plan being edited can be used twice or more in a series of dialogues. By setting this item to "possible", this plan can appear more than once in a series of dialogues. A reusable dialogue plan is a question for which there are multiple possible answers, such as "Which other countries have you traveled to?" In the case of a dialogue plan that cannot be reused, the "reusability" is set to "no".

"Score" is a score that defines the probability of selection of this dialogue plan. The example shown in FIG. 9 is a dialog plan that can also be used as an initial utterance and can be started without prior knowledge, so the score is a constant. In the case of an interaction plan whose selection depends on the interaction that has already taken place, a formula is entered in which the score varies according to the parameters of the interaction that has taken place so far.

"Belonging group" designates the group to which the dialog plan being edited belongs. A belonging group is a group that can be collectively designated as a transition destination. "Person in charge" is the person in charge who generated this dialogue plan, and "Status" is a diagram showing the editing status of this dialogue plan. As described above, in the present embodiment, there are two statuses of "undetermined" and "determined", and in the example shown in FIG. 9, it is "undetermined". Editing a plan automatically changes the status to "unconfirmed". In other words, when creating a new plan, the status becomes "unconfirmed", and even if the plan was confirmed, if you edit it on the edit screen, the status of the interactive plan will change to "unconfirmed". Updated. "Last update date" is the date and time when this interaction plan was last updated.

In the column of "pretending utterance", the pretending utterance of this dialogue plan is displayed, but in the example shown in FIG. 9, there is no pretending utterance created yet, and "new registration" is displayed. When you click "New registration", the screen changes to the swinging utterance editing screen.

In the "user response→received utterance" column, the user's response to the shaking utterance and the received utterance corresponding thereto are displayed. In the example shown in FIG. 9, an "exceptional pattern" and "silent utterance" are displayed by default, although the swinging utterance has not yet been created. Clicking "new registration of user response" transitions to a user response edit screen, and clicking "new registration of received utterance" transitions to a received utterance edit screen.

FIG. 10 is a diagram showing an example of a screen for editing the swinging utterance. "Plan name" is the name of the dialog plan being edited. The "Update" button at the top of the swinging utterance editing screen has a function of saving the latest content edited on the swinging utterance editing screen.

"Summary" in the column of "details of utterance" is a specific description of the content of the dialogue plan, and is the same as the "plan outline" on the dialogue plan editing screen. The "conditions" include "previous speaker", "immediate dialogue act classification", and "immediate dialogue act", and the conditions for the dialogue act immediately before the swinging utterance can be set. "Dialogue act" is an item for setting what kind of dialogue act the swinging utterance is classified into.In the example shown in FIG. 10, it is classified as a personal information question.

The "response format" is the content of the gesture speech, and the "response example" is an example of the gesture speech made according to the response format. In the examples shown in FIG. 10, all of the sentences are "Do you have a favorite food?" In this case, by having the "response example", it is possible to grasp what kind of utterance the pretended utterance described in the "response format" was intended to be created.

A “query sentence” is a query sentence used when retrieving this pretended utterance, and a “query index” is a parameter included in the “query sentence”. A swing utterance score is calculated based on whether or not it corresponds to the query sentence, and a swing utterance is selected based on the calculated score.

FIG. 11 is a diagram showing an example screen for editing a user response. At the top of the user response editing screen, there are buttons for "Update", "Create Copy" and "Delete". The "update" button has a function of saving the latest content edited on the user response editing screen. A "delete" button is a button for deleting this interaction plan. The "copy creation" button will be described later. "Plan Name" is the name of the dialogue plan being edited, and "Pretended Utterance" describes the contents of the pretended utterance created on the pretended utterance edit screen.

"Summary" in the "user utterance details" column is a summary of the user response. In the example shown in FIG. 11, the user response is to "Do you have a favorite food?" The "conditions" include "previous speaker", "immediate dialogue act classification", and "immediate dialogue act", and it is possible to set the conditions for the dialogue act immediately before the user response.

A "template" is an item for creating or designating responses that are commonly used for some swaying utterances. Here, the user response data for the designated template is not directly described in this interaction plan, but is stored in an external storage unit. The interaction plan references external user response data. Therefore, when the template data is rewritten, the change is reflected in all interaction plans that refer to that template.

"Existing Selection" of "Template" is a button for selecting an already created template. In the example shown in FIG. 11, the template of "utterance including dish name-extended named entity" is selected. "Overwrite" is a button for overwriting an existing template. If you select "overwrite", you can change other dialogue plans that refer to the template at once. "New" is a button for creating a new template.

The “dialogue action” is an item representing the classification of the dialogue action of the user response. In the example shown in FIG. 11, they are classified as answers. "Utterance slot pattern" is a pattern for detecting words included in a predetermined category included in user responses. In the example shown in FIG. 11, "Food" and "Dish" are described, and the slot patterns are food and dish names. A "speech regular expression pattern" is a pattern of strings that a user response is expected to contain. In the example of FIG. 11, no utterance regular expression pattern is described. As an example of the utterance regular expression pattern, the regular expression for the user response that the price was not high is (normal|normal|fair|bad|bad|sorry|not so much|a?mary|so)*( Expressions such as "expensive | cheap" are conceivable. An utterance example is an example of utterance corresponding to the utterance slot pattern and the utterance regular expression pattern.

The column "list of received utterances" displays a list of received utterances following the user response. In the example of FIG. 11, the "summary" is linked to the list of received utterances of "response to answering the name of the dish" (the check box of the link is checked). Here, the "number of links" indicating how many user responses are linked to the received utterance of "response to answering the name of the dish" is described. As a result, it can be determined whether the received utterance "response to answering the name of the dish" is a frequently used received utterance or a rarely used received utterance.

Next, the check function of the user response edit screen will be described. On the user response edit screen, when the "update" button is pressed, the contents are checked before saving the edited contents of the user response. Specifically, the "example response" is referenced to determine whether the "utterance slot pattern" and "utterance regular expression pattern" match the "example response".

FIG. 12 is a diagram showing an example screen for editing a user response. In the screen shown in FIG. 11, the screen shown in FIG. It is a figure which shows the state which described as "Japanese food" and pushed the "update" button. In this case, the utterance example does not match the dish name described in the "utterance regular expression pattern" (because the utterance regular expression pattern includes "Japanese food" and "Japanese food" but not "Japanese food"). , displays an input value error message "Example utterance" does not match the utterance regular expression pattern.

Since the example utterance describes a typical example of a user response assumed by the worker, the utterance slot pattern is determined by determining whether or not the utterance slot pattern and the utterance regular expression pattern match the example utterance. You can check whether the "utterance regular expression pattern" is correct.

FIG. 13 is a diagram showing an example of a received utterance editing screen. At the top of the received utterance editing screen, there are buttons for "update", "copy creation" and "delete". The "update" button has a function of saving the latest content edited on the user response editing screen. A "delete" button is a button for deleting this interaction plan. The "copy creation" button will be described later. "Plan Name" is the name of the dialogue plan being edited, and "Pretended Utterance" describes the contents of the pretended utterance created on the pretended utterance editing screen.

"Summary" in the "Details of Received Utterance" column is an explanation of the contents of the "Received Utterance". Conditions can be set for the last interactive action to be taken. The “dialogue act” is an item representing the classification of the dialogue act of the received utterance. In the example shown in FIG. 13, it is classified into impression/opinion response.

The "response format" is the content of the received utterance, and the "response example" is an example of the uttered utterance made according to the response format. In the example shown in FIG. 13, the "response format" is "I like <<<o1>>>." “o1” in the response example is the name of the dish included in the user response. As a result, in the example shown in FIG. 13, the dish name included in the user response is parroted. An example of "response" is "I like ramen".

A “query sentence” is a query sentence used when retrieving this pretended utterance, and a “query index” is a parameter included in the “query sentence”. In the example shown in FIG. 13, it is a condition that the liking of a certain dish is grounded.

"Transition plan ID" designates the next dialogue plan to transition from the dialogue plan being edited. There is a display that says "Choose from a list", and a link is pasted. When "select from list" is selected, a list for selecting a transition destination shown in FIG. 14 is displayed. This list contains information such as "plan name", "plan overview", availability of initial utterance (described as "first"), availability of reuse (described as "re-"), status, and person in charge. include. By checking the "Select" column at the left end of this list, a dialog model in this list can be selected as a transition destination.

Returning to FIG. 13, the ``transition destination plan group ID'' designates the next dialogue plan to which the dialogue plan being edited is to be transitioned. ID". "Transition destination plan group ID" also has a display of "select from list" and a link is attached. When "select from list" is selected, a list for selecting a transition destination group shown in FIG. 15(a) is displayed. The transition destination group describes the group name of the dialogue plan group. By selecting this group name, all dialogue plans included in that group can be selected as transition destinations. It is efficient because it is not necessary to select the dialogue plan one by one.

FIG. 15B is a diagram showing an example of a screen for grouping dialogue plans. When creating a new group, select "Add Group" and "Assign Plan" to add new interaction plans to the group.

Next, an execution example of the copy function will be described with reference to FIGS. 16 to 18. FIG. FIG. 16 is an example of the plan edit screen of the copy source interaction plan. In the "user response→received utterance" column at the bottom of the screen shown in FIG. 16, there is a "copy selection" check box at the left end of the list of user responses and received utterances. By checking this check box and selecting the "Create copy" button at the top of the dialogue plan edit screen, the corresponding user response and received utterance are saved in the temporary storage area (clipboard). .

FIG. 17 is an example of the plan edit screen of the interaction plan to be copied. When the user response and the received speech are stored in the temporary storage area, the "Paste User Response" button becomes selectable. FIG. 17 shows an example in which the "Paste User Response" button can be selected. On the screen shown in FIG. 17, when the "Paste User Response" button is selected, the user response and received utterance checked on the screen of FIG. added to the column. By copying a user response and a received utterance created in one dialogue plan to another dialogue plan in this way, the user response and the received utterance can be easily generated.

(test function)
FIG. 19 is a diagram showing an example of a dialogue plan test screen. By clicking "to plan test" displayed at the top of the dialogue plan edit screen, the dialogue plan test screen shown in FIG. 19 is displayed. At the top of the dialog plan test screen, there are buttons for "reset dialog", "select data source", and "update profile". "Dialogue reset" has the function of resetting the dialog being executed in the plan test. Used to run the test of the dialog plan from scratch. "Select Data Source" is a button for selecting a data source to be used when executing an interaction plan. For example, when testing a conversation plan about baseball, the team, players, game results, etc. are acquired from the data source and used for the conversation test. The "update profile" button is a button for updating the user's profile. The user's profile is updated because the selected topic changes according to the user's age, gender, taste, and the like.

In FIG. 19, an interaction plan on the topic of baseball is shown being tested. Based on the data acquired from the data source called Data Stadium, the dialogue system utters a pretended utterance, "Did you watch the match between Team A and Team B four days ago?" On the right side of the test screen, information is displayed in the "system utterance details" and "foundation information" columns. Foundational information is generated based on the data obtained from the data sources and the user's profile.

At the bottom of the plan test screen, there is a field for entering a user response, labeled "Please enter a user response." You can freely write your user response in the box below the text "Enter User Response". In the user response entry field, a list of user responses prepared by the interaction plan is displayed, and the user response to be tested may be selected from among these.

FIG. 20 is a diagram showing an example of a test screen when a user response of "Who threw the ball?" is input on the test screen shown in FIG. The user response "Who threw?" does not correspond to the adjacent triple pattern, as shown in the "User response details" column, "The rule type returned a non-adjacent triple pattern." The user response returned here is "DQ: Dynamic Question Answering Pattern", and the system retrieves the starting pitcher's data from the data source, generates system utterances based on the retrieved data, and tests Display at the bottom of the screen. Here, if the system utterance "A team pitched by A, and B team threw by B.", the system utterance is displayed on the test screen as shown in FIG. The configuration of the screen in FIG. 21 is the same as the screen shown in FIG.

In the screen of FIG. 21, the interaction of the worker saying "That's right" and "Team A pitched it, and Team B pitched it to Team B." Therefore, the user response "that's right" corresponds to the second user response of the adjacent triple pattern. When a user response of "that's right" is entered, the test screen transitions to FIG. As shown in FIG. 22, the received utterance from the system, "You are in good shape when you are winning" was returned, and it was confirmed that the dialogue was established.

FIG. 23 is a diagram showing an example of a screen displaying the results of the automatic test of the interaction plan. Although FIGS. 19-22 show examples of manual testing in which the operator can manually enter arbitrary user responses, testing can also be performed automatically. "Plan ID" displays the plan ID of the interactive plan to be tested. When the test execution button is moved to the right, the test is executed, and "test No.", "matching result", "assumed pattern ID", and "user response sentence" are listed on the user response pattern test screen. The matching result is OK when the user response matches the expected pattern ID, and the pattern is wrong when they do not match. Also, when an exception pattern matches as a user response, "exception" is displayed. At the top of the list, the test results are displayed numerically as "OK rate", "pattern difference rate", "exception rate", and "error rate".

What has been described above is the function of testing whether or not the utterance utterance, the user response, and the utterance received match in each dialogue plan. , and has a function to test whether the dialogue continues. This test is called a "batch test". A batch test function is executed in the dialogue plan creation support device 10 .

FIG. 24 is a diagram showing an example of a batch test screen. The batch test screen has a "test execution" button and a "status update" button. A "test execution" button is a button for executing a test. Moving the button to the right executes the test, and moving the button to the left interrupts the test being executed.

A batch test is a test using a dialogue plan stored in the dialogue plan DB 14 of the dialogue system support device. A predetermined dialogue plan is specified, transitions are made from that dialogue plan using dialogue plans stored in the dialogue plan DB 14, and the number of dialogue plan transitions in which the dialogue continues is counted. If the dialogue continues for a long time and ends with the closing utterance, it is considered successful, and if there is no transition destination and the dialogue is immediately interrupted, it cannot be said to be successful. Here, to supplement the closing utterance, it is assumed that a dialogue plan of "closing utterance" for ending the conversation is prepared in the dialogue plan. The “closing utterance” dialogue plan is a dialogue plan in which “E: closing utterance pattern” is set in the “rule type”. It should be noted that the closing utterance may not be set for the dialogue plan, but may be set for the receiving utterance included in the dialogue plan. As a result, even in one dialogue plan, detailed control can be performed such that when a certain received utterance is made, the dialogue is terminated, and when another received utterance is made, the dialogue plan is changed. As a method of setting the end of the dialogue for the received utterance, for example, in the received utterance of the dialogue plan, it is possible to set the transition destination to "end".

Also, even if the dialogue continues, it may go through an exception pattern on the way. Exceptional patterns are user response patterns that can respond to any type of interaction, such as backtracking and listening. It also counts the number of passes.

The "status update" button is a button for selecting whether or not to display the updated test results on the batch test screen, and moving the button to the right displays the updated test results. During the execution of the test, the test results are updated along with the progress. Therefore, if you want to check the test results up to a certain point during the execution of the test, move the "update status" button to the left to stop updating the test results.

In the "Summary" column, the test results displayed in the "Total by plan" column below are aggregated and displayed. The "number of final utterances reached", "number of passages through exceptions", "number of no transitions", etc. indicate the test results.

FIG. 25(a) is a graph showing the number of final utterance arrivals. The vertical axis indicates the number of times the concluding utterance is reached, and the horizontal axis indicates the number of dialogue plan transitions made until the concluding utterance is reached. Based on this result, it is possible to determine whether the dialogue plan is transitioning successfully.

FIG. 25(b) is a graph showing the results of the exception transit count. The vertical axis shows the number of exception patterns passed through, and the horizontal axis shows the number of dialogue plan transitions until reaching the closing utterance. This makes it possible to analyze how many of the trials that reach the closing utterance go through the exception pattern.

FIG. 25(c) is a graph showing the results of the no-transition count. The vertical axis shows the number of times the transition destination was missed, and the horizontal axis shows how many transitions it occurred after. From this result, it can be judged that there is a problem in the method of specifying the transition destination of the dialogue plan if there is no transition destination at an early stage.

FIG. 26 is a diagram showing an example of hierarchical display of a plan list in the dialogue plan creation support device 10 of the present embodiment. As shown in FIG. 26, by hierarchically displaying the transition relationships of the plans, the spread and flow of the dialogue can be grasped. In addition, hierarchical display makes it easier to understand that the dialogue plan is looping. A warning may be displayed if there is a loop in the interactive plan when displayed hierarchically. FIG. 26 shows an example of displaying a warning "'Bread or Rice?_1' is repeated." Also, when displaying the warning, the number of times to return to the same interaction plan may be displayed. There is a problem if it returns to the same dialogue plan in a short cycle, but if the cycle until it returns to the same dialogue plan is long, it may not be a problem. By indicating the number of times, it is possible to easily judge whether the loop is permissible or not.

INDUSTRIAL APPLICABILITY The present invention is useful, for example, as a device that assists a plurality of workers in creating a dialogue plan.

1 dialogue plan creation support device 11 operation unit 12 control unit 13 communication unit 14 dialogue plan DB
20 work terminal

Claims (14)

A dialogue plan consisting of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user, which includes information on the transition destination of the next dialogue plan. A device for assisting creation,
an interaction plan database that stores the interaction plan;
a communication unit that communicates with a worker terminal in which the worker creates and inputs the interaction plan;
Transmitting, to the worker terminal via the communication unit, a dialog plan creation support screen displaying candidates for a dialog plan to be transitioned to , and receiving data of the dialog plan input on the creation support screen. a control unit that stores the data of the dialog plan obtained in the dialog plan database;
with
The creation support screen displays the number of backlinks to which a candidate for a transition destination dialogue plan is linked as a transition destination . The control unit gives an undetermined status to the dialogue plan when storing data of the dialogue plan received from the worker terminal, and changes the undetermined status to finalized according to an operation by an administrator. 2. The dialogue plan creation support device according to claim 1 , wherein the dialogue plan creation support device is modified. 3. A dialog plan creation support device according to claim 1, wherein candidates for transition destination dialog plans are grouped on said creation support screen. A dialogue plan consisting of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user, which includes information on the transition destination of the next dialogue plan. A device for assisting creation,
an interaction plan database that stores the interaction plan;
a communication unit that communicates with a worker terminal in which the worker creates and inputs the interaction plan;
Transmitting an interaction plan creation support screen to the worker terminal via the communication unit, receiving interaction plan data input on the creation support screen, and storing the received interaction plan data in the interaction plan database a control unit that stores
with
The control unit outputs a warning when there is a looping portion in the transition of the dialogue plan,
The said control part outputs the frequency|count which shows how many times it returns to the same dialogue plan . A dialogue plan consisting of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user, which includes information on the transition destination of the next dialogue plan. A device for assisting creation,
an interaction plan database that stores the interaction plan;
a communication unit that communicates with a worker terminal in which the worker creates and inputs the interaction plan;
Transmitting an interaction plan creation support screen to the worker terminal via the communication unit, receiving interaction plan data input on the creation support screen, and storing the received interaction plan data in the interaction plan database a control unit that stores
with
The control unit starts from one selected dialogue plan and uses the dialogue plan stored in the dialogue plan database to simulate the dialogue until the closing utterance is reached or there are no transition destinations, and then continue. A dialog plan creation support device that counts the number of dialog plans created and displays the result on a display unit . 6. The dialogue plan creation support device according to claim 5 , wherein the control unit counts the number of times each dialogue plan is used and displays the result on the display unit. 7. The dialogue plan creation support device according to claim 5 or 6 , wherein the control unit causes the display unit to display, as reasons for ending the dialogue, the number of times a final utterance was reached and the number of times a transition destination was lost. the interaction plan includes general user responses that can be used regardless of the content of the utterance;
8. The dialogue plan creation support device according to any one of claims 5 to 7 , wherein the control unit causes the display unit to display the number of times that a general user response is matched. A dialogue plan consisting of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user, which includes information on the transition destination of the next dialogue plan. A method of assisting in the creation of
a step of transmitting, to the worker terminal where the worker creates and inputs the dialog plan, a dialog plan creation support screen displaying candidates for the dialog plan to be transitioned to ;
a step of accepting input of dialogue plan data through the creation support screen at the worker terminal;
a step of receiving dialogue plan data input at the worker terminal and storing the received dialogue plan data in a dialogue plan database;
with
A dialog plan creation support method in which the creation support screen displays the number of backlinks to which candidates for a transition destination dialog plan are linked as transition destinations . A dialogue plan consisting of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user, which includes information on the transition destination of the next dialogue plan. A method of assisting in the creation of
a step of transmitting a dialog plan creation support screen to a worker terminal where the worker creates and inputs the dialog plan;
a step of accepting input of dialogue plan data through the creation support screen at the worker terminal;
a step of receiving dialogue plan data input at the worker terminal and storing the received dialogue plan data in a dialogue plan database;
outputting a warning if there is a loop in the transition of the dialog plan;
outputting a count indicating how many times to return to the same dialogue plan;
A dialogue plan creation support method comprising: A dialogue plan consisting of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user, which includes information on the transition destination of the next dialogue plan. A method of assisting in the creation of
a step of transmitting a dialog plan creation support screen to a worker terminal where the worker creates and inputs the dialog plan;
a step of accepting input of dialogue plan data through the creation support screen at the worker terminal;
a step of receiving dialogue plan data input at the worker terminal and storing the received dialogue plan data in a dialogue plan database;
Using one selected dialogue plan as a starting point and using the dialogue plans stored in the dialogue plan database, dialogue simulation is performed until the final utterance is reached or there are no transition destinations, and the number of continued dialogue plans a step of counting and displaying the result on the display unit;
A dialogue plan creation support method comprising: A dialogue plan consisting of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user, which includes information on the transition destination of the next dialogue plan. A program for assisting in the creation of a computer,
a step of transmitting, to the worker terminal where the worker creates and inputs the dialog plan, a dialog plan creation support screen displaying candidates for the dialog plan to be transitioned to ;
a step of accepting input of dialogue plan data through the creation support screen at the worker terminal;
a step of receiving dialogue plan data input at the worker terminal and storing the received dialogue plan data in a dialogue plan database;
and
The creation support screen is a program for displaying the number of backlinks to which dialogue plan candidates of transition destinations are linked as transition destinations . A dialogue plan consisting of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user, which includes information on the transition destination of the next dialogue plan. A program for assisting in the creation of a computer,
a step of transmitting a dialog plan creation support screen to a worker terminal where the worker creates and inputs the dialog plan;
a step of accepting input of dialogue plan data through the creation support screen at the worker terminal;
a step of receiving dialogue plan data input at the worker terminal and storing the received dialogue plan data in a dialogue plan database;
outputting a warning if there is a loop in the transition of the dialog plan;
outputting a count indicating how many times to return to the same dialogue plan;
program to run. A dialogue plan consisting of a gesture utterance from the system to the user, a user response from the user to the system, and a received utterance from the system to the user, which includes information on the transition destination of the next dialogue plan. A program for assisting in the creation of a computer,
a step of transmitting a dialog plan creation support screen to a worker terminal where the worker creates and inputs the dialog plan;
a step of accepting input of dialogue plan data through the creation support screen at the worker terminal;
a step of receiving dialogue plan data input at the worker terminal and storing the received dialogue plan data in a dialogue plan database;
Using one selected dialogue plan as a starting point and using the dialogue plans stored in the dialogue plan database, dialogue simulation is performed until the final utterance is reached or there are no transition destinations, and the number of continued dialogue plans a step of counting and displaying the result on the display unit;
program to run.

Images