Device, generation device, and method

Abstract

To provide a device and a method that can improve a user's completion probability.SOLUTION: An interactive device 100 of the present disclosure includes an answer table 103 (storage unit) that stores answer information (action information) including a procedure content (final action) and recommended items (additional actions) related to the procedure content. The interactive unit 101 functions as an acquisition unit that acquires user inquiries and an answer unit. In response to reception of an inquiry from a user terminal 200, the interactive unit refers to the answer table 103 and answers about the procedure content (final action) corresponding to the inquiry and the recommended items (additional actions) corresponding to the procedure content.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present disclosure relates to an apparatus, a generating apparatus and a method for assisting user behavior. [Background technology]

[0002] Patent Document 1 describes a question-answering system that provides an interactive UI. This question-answering system has a function of estimating a user's feelings toward a query and generating training data that includes corresponding emotional expressions. The training data generation system adds emotional information in an emotional expression unit, and transmits the data to an answer generation engine in a training data providing unit. The UI server receives the user's query, sends it to the answer generation engine, receives the answer including the emotional expression, and transmits it to the user. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2019-117517 Summary of the Invention [Problem to be solved by the invention]

[0004] In order to provide detailed customer support tailored to each individual customer, companies need to understand the needs and issues from customer information such as attributes and behavior, and provide support in the most optimal way.

[0005] The technology described in Patent Document 1 cannot respond to customers whose requests are unclear, and there are cases where the user's request cannot be completed.

[0006] In order to solve the above problem, the present disclosure aims to provide an apparatus, a generation apparatus, and a method that can improve a user's completion probability. [Means for solving the problem]

[0007] The device of the present disclosure includes a memory unit that stores answer information including a final action and an additional action related to the final action, an acquisition unit that acquires a user's inquiry, and an answer unit that, upon receiving the inquiry, refers to the memory unit and answers the inquiry with a final action and an additional action corresponding to the final action. [Effects of the Invention]

[0008] According to the present disclosure, the probability of a user completing a final action can be improved. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram showing a system configuration including a dialogue device 100 that functions as a chatbot according to the present disclosure. [Figure 2] FIG. 2 is a block diagram showing the functional configuration of the dialogue device 100 of the present disclosure. [Figure 3] FIG. 3 shows a specific example of the answer table 103 and an outline of its processing. [Figure 4] FIG. 4 is a flowchart showing the operation of the interactive device 100 of the present disclosure. [Figure 5] FIG. 5 is a display screen showing a dialogue with the dialogue device 100 on the display screen of the user terminal 200. As shown in FIG. [Figure 6] FIG. 6 is a block diagram showing the functional configuration of the answer table generating device 300 for generating the answer table 103 of the present disclosure. [Figure 7] FIG. 7 is a flowchart showing the operation of the response table generating device 300. [Figure 8] FIG. 8 is a diagram showing a typical process for generating an answer table. [Figure 9] FIG. 9 is a diagram showing an example of a process for calculating the degree of association between a sub-action and a final action. [Figure 10] FIG. 10 is a diagram showing the learning process of a large-scale behavior model. [Figure 11] FIG. 11 is a diagram illustrating an example of a process for calculating the difference in completion probability. [Figure 12] 1 is a diagram illustrating an example of the hardware configuration of a dialogue device 100 and a response table generating device 300 according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0010] The present disclosure will be described with reference to the accompanying drawings. Whenever possible, the same parts are designated by the same reference numerals and redundant description will be omitted.

[0011] 1 is a diagram showing a system configuration including an interactive device 100 that functions as a chatbot according to the present disclosure. As shown in the figure, the interactive device 100 can be connected to a user terminal 200 via a network for communication. The user terminal 200 transmits text such as a question to the interactive device 100, and the interactive device 100 transmits an answer to the question indicated in the text.

[0012] 2 is a block diagram showing the functional configuration of the dialogue device 100 of the present disclosure. As shown in the figure, the dialogue device 100 includes a dialogue unit 101, a query content acquisition unit 102, and a response table 103.

[0013] The dialogue unit 101 is a part that dialogues with the user terminal 200, receives text information transmitted from the user terminal 200, and transmits text information to the user terminal 200. Note that in the present disclosure, the description is given on the premise that the user and the dialogue device 100 dialogue by transmitting and receiving text information, but this is not limited to this, and the dialogue may also be performed using image information or audio information.

[0014] The inquiry content acquisition unit 102 is a part that acquires (understands) the content of the user's inquiry by analyzing the text (question) acquired by the dialogue unit 101 using a known language model. Then, the inquiry content acquisition unit 102 refers to the answer table 103 to acquire the procedure content and recommended items that are the answer corresponding to the inquiry content, and the dialogue unit 101 transmits the procedure content and recommended items to the user terminal 200.

[0015] The answer table 103 is a storage unit that stores the inquiry content, the answer, the recommended items, and the effect for each user. FIG. 3 is a diagram showing a specific example of the answer table 103 and an overview of its processing. As shown in the figure, the answer table 103 stores, for a user ID of AAA, the inquiry content: model change, the answer: model change procedure, the recommended items: model simulation, and the effect: +50%. The answer corresponds to the final action in response to the inquiry content. The recommended items correspond to additional actions in response to the answer. In the present disclosure, the answer table 103 associates different recommended items and effects for each user, but this is not limiting, and the same recommended items and effects may be associated with all users.

[0016] The inquiry content acquisition unit 102 compares the user ID: AAA and inquiry content: model change as information obtained from the dialogue, and as a result, returns the following information to be included in the dialogue: answer: model change procedure (including information or links to carry out the procedure) and recommended items: model simulation.

[0017] In the present disclosure, the inquiry content indicates a request from a user, and the answer indicates a response to the request. The answer is information for carrying out the request and for causing the user to take a final action. The recommended items are additional items related to the answer. The effect is the effect of adding the recommended items and indicates the probability that the user will achieve the procedure indicated in the answer.

[0018] The operation of the dialogue device 100 configured as above will be described. Fig. 4 is a flowchart showing the operation of the dialogue device 100 of the present disclosure. When text information is transmitted as a dialogue from the user terminal 200, the dialogue unit 101 acquires the user ID and the text information, and identifies the user based on the user ID (S101).

[0019] The inquiry content acquisition unit 102 acquires the inquiry content from the text information acquired by the dialogue unit 101 (S102). The inquiry content acquisition unit 102 refers to the answer table 103 and determines whether there is an answer and a recommended item corresponding to the user ID and the inquiry content (S103). When multiple recommended items such as a first recommended item and a second recommended item are stored in the answer table 103, the inquiry content acquisition unit 102 further acquires one recommended item based on the effect associated with that item. Note that multiple recommended items may be acquired. For example, the recommended item with the higher effect is acquired.

[0020] If the result of step S103 is YES, the query content acquisition unit 102 acquires an answer and recommended items corresponding to the query content (S104). If the result of step S103 is NO, the query content acquisition unit 102 acquires an answer (S104).

[0021] The dialogue unit 101 generates text information including the answer (and the recommended items) and transmits it to the user terminal 200 (S106). This text information is generated as dialogue-style sentences.

[0022] The recommended items in the present disclosure are information related to an answer to an inquiry. If the answer to the inquiry is a procedure for a certain purpose, the recommended items are items that indicate (convenient) procedures that should be recommended for the user to perform the procedure for the purpose.

[0023] In this way, the interactive device 100 transmits an answer to the query and recommended items to the user terminal 200, so that the user can obtain an answer to the query and also obtain recommended items to better execute the query.

[0024] FIG. 5 shows a display screen of the user terminal 200 displaying a dialogue with the dialogue device 100. FIG. 5(a) shows an answer that does not include a recommended item. FIG. 5(b) shows an answer that includes a recommended item. Answer K1 is an answer to the inquiry. Answer K2 indicates a recommended item. Anchors are embedded in answers K1 and K2. By tapping answers K1 and K2, the user transitions to a specific screen, where the user can perform the action requested in the inquiry. In the figure, answer K1 indicates transition to a procedure screen for upgrading the mobile phone model. Answer K2 is a recommended item and indicates a link destination for transitioning to a screen that is convenient for upgrading the model. By tapping answer K2, the user transitions to that screen and can perform a model simulation. The model simulation is, for example, a screen that allows the user to actually operate the functions of the mobile phone. Through this model simulation, the user can proceed with the model upgrade procedure, which ultimately increases the probability of completing the action of upgrading the model.

[0025] In this way, the answer table 103 stores answers to inquiries in association with their recommended items, and the inquiry content acquisition unit 102 can acquire the recommended items along with the answers to the inquiries by referring to the answer table 103. The answer table 103 also stores the effects of the recommended items. When there are multiple answers corresponding to the inquiry content, the inquiry content acquisition unit 102 may acquire one answer by referring to the effects. Furthermore, the dialogue unit 101 may generate text information to be sent to the user so as to list the recommended items in descending order of effectiveness.

[0026] Next, we will explain how to generate this answer table 103. The answer table 103 is prepared in advance by the operator of the dialogue system, and may be created manually by the operator, or, as will be explained below, efficient recommended items and their effects may be derived based on each user's behavior (web browsing, conversation with a chatbot, real-life behavior), and stored in the answer table 103.

[0027] 6 is a block diagram showing the functional configuration of the answer table generating device 300 for generating the answer table 103 of the present disclosure. In the following description, the answer above will be referred to as a final action because it encourages the user to take a final action (procedure). The recommended items encourage additional actions to the final action and are important actions, so they will be referred to as important actions.

[0028] As shown in the figure, the answer table generation device 300 includes a learning data acquisition unit 21, a classification unit 22, an output model 23, an extraction unit 24, an important behavior acquisition unit 25, a model generation unit 26, a completion probability estimation model 26a, a completion probability acquisition unit 27, a calculation unit 28, a judgment unit 29, and an answer table generation unit 30. This answer table generation device 300 is a device for generating an answer table 103, and can generate inquiry contents and corresponding recommendation items by learning the behaviors of multiple users. The operation of each component will be explained below using the flowchart in FIG. 7. FIG. 7 is a flowchart showing the operation of this answer table generation device 300.

[0029] The learning data acquisition unit 21 acquires a predetermined number of user actions (or a predetermined period) (step S21). In step S21, the learning data acquisition unit 21 refers to the user action DB 400 and acquires the type of user action, the content of the user action, and the step number for each user.

[0030] This user behavior DB 400 is a database that reflects user behavior, including interactions with a chatbot (dialogue system), browsing of web pages, call center response history, and behavior when actually visiting a store. In particular, interactions with a chatbot or browsing of web pages are performed via a portal server (not shown), and are therefore registered by the portal server. The user terminal 200 accesses the chatbot or browses web pages on the Internet via the portal server. Behavior when actually visiting a store is registered in the user behavior DB 400 by a store clerk.

[0031] In the present disclosure, it is assumed that a user performs actions such as changing fees and changing models on a mobile phone shop site.

[0032] Next, the classification unit 22 classifies the users into either a first user group or a second user group for each final action (step S22). For example, if the final action is a change of model, the users are classified into a first user group who have changed the model (reached the final action) and a second user group who have not changed the model (reached the final action).

[0033] In the present disclosure, a user accesses a mobile phone shop site via user terminal 200 and performs a final action. The final action refers to the final action taken to achieve a desired goal. In the present disclosure, the final action is the action of confirming a procedure, such as a fee change procedure or a model change procedure, on the mobile phone shop site, and is the action of pressing a change confirmation button.

[0034] In step S22, the classification unit 22 determines whether the user's step number is the final step number (in the above example, the final step number corresponding to changing models) based on the step number acquired for each user in step S21. In step S22, if the classification unit 22 determines that the user's step number is the final step number, the user is classified into a first user group. If the classification unit 22 determines that the user's step number is not the final step number, the user is classified into a second user group.

[0035] Next, the output model 23 outputs the degree of association between one or more sub-actions acquired in step S21 and the final action (step S23). In step S23, the output model 23 uses a known means to output the degree of association between the sub-actions and the final action. The output model 23 will be described later.

[0036] Next, the extraction unit 24 extracts a predetermined number (or a predetermined period) of small actions of the first user group and a predetermined number (or a predetermined period) of small actions of the second user group based on the degree of association output in step S23 (step S24). In the present disclosure, in step S24, the extraction unit 24 extracts a predetermined number of small actions with a high degree of association (above a threshold) from among the small actions performed by users of the first user group. Also, in step S24, the extraction unit 24 extracts a predetermined number of small actions with a high degree of association (above a threshold) from among the small actions performed by users of the second user group.

[0037] Instead of extracting small actions with high relevance, a candidate table for storing small action candidates may be prepared, and the extraction unit 24 may extract the small actions from the candidate table and calculate the completion probability for each of them in the subsequent processing (step S25 onwards). This candidate table is assumed to be generated in advance by the operator of the response table generation device 300.

[0038] Next, the important action acquisition unit 25 acquires an important action from one or more small actions acquired in step S24 based on the degree of association output in step S23 (step S25). In the present disclosure, in step S25, the important action acquisition unit 25 acquires, as an important action, a small action that is among the predetermined number of small actions of the first user group extracted in step S24 but is not among the predetermined number of small actions of the second user group. Note that a specific small action may be determined as an important action in advance by the analyst.

[0039] Next, the model generation unit 26 generates a completion probability estimation model 26a by machine learning using the predetermined number (predetermined period) of user actions acquired in step S21 as explanatory variables and whether or not the final action was reached as a target variable (step S26). The completion probability estimation model 26a generated in step S26 estimates the completion probability of reaching the user's final action (for example, a procedure for changing model, etc.) by inputting the predetermined number (or predetermined period) of user actions of a specific user (for example, user AAA). do.

[0040] Specifically, the model generation unit 26 generates a completion probability estimation model by setting a model in which, for example, user behavior is used as an explanatory variable and whether the user has reached the final behavior is used as a target variable, and optimizing the coefficients (parameters) of the model through learning. The model may be a linear model such as logistic regression, or a nonlinear model such as a multi-layer neural network. The model may be any model that can set a one-dimensional output for a multi-dimensional input.

[0041] For example, the explanatory variables may be generated by applying a technology used in natural language processing, such as BERT (Bidirectional Encoder Representations from Transformers). As an example, first, each user action is defined as a token, and a model is trained that uses a sequence of these as input and outputs a semantic vector for the token at a certain point in time. Then, the user action is input to the model, and the output semantic vector at the time of the action may be used as an explanatory variable for the completion probability estimation model. The objective variable may be created by setting the value to 1 when a user's step number is the final step number and 0 when the user's step number is not the final step number.

[0042] As an example, the completion probability estimation model 26a receives each user action as input. The completion probability estimation model inputs each user action into a BERT or the like, and outputs a semantic vector for each user action. The completion probability estimation model 26a inputs the output semantic vector into a model such as logistic regression, and outputs a completion probability. However, the processing content of the completion probability estimation model 26a can be changed as appropriate.

[0043] Next, the completion probability acquisition unit 27 uses the completion probability estimation model 26a generated in step S26 to acquire, for each user, a completion probability of reaching a final action of the user action not including the significant action (step S27). In step S27, the completion probability acquisition unit 27 inputs the user action into the completion probability estimation model as a user action not including the significant action. In the present disclosure, actions in a period not including the significant action may be selected from the user's actions in a predetermined period, or actions may be generated by excluding the significant action from the user's actions in a predetermined period.

[0044] In step S27, the completion probability acquisition unit 27 uses the completion probability estimation model 26a generated in step S26 to acquire, for each user, a completion probability of the user action including the significant action reaching a final action. In step S27, the completion probability acquisition unit 27 inputs the user action to which the significant action acquired in step S25 has been added as a user action including the significant action into the completion probability estimation model. In step S27, for each significant action acquired in step S25, a completion probability of the user action including the significant action is acquired.

[0045] In the present disclosure, a "user behavior to which an important behavior has been added" refers to a user behavior in which the timing (date and time) at which the important behavior will be performed is predicted, and the important behavior is added to the user behavior by replacing other behavior at that timing (date and time). The prediction of the timing at which the important behavior will be performed will be described later. Note that the timing of the important behavior may be determined in advance, and other behavior at that timing may be replaced with the important behavior.

[0046] Next, the calculation unit 28 calculates the difference between the completion probability of the user behavior including the important behavior acquired in step S27 and the completion probability of the user behavior not including the important behavior (step S28). In step S28, the calculation unit 28 calculates the difference for each of the important behaviors acquired in step S25.

[0047] Next, the determination unit 29 determines whether or not the important action is effective based on the difference calculated in step S28 (step S29). In step S29, the determination unit 29 calculates the ratio of the number of users whose lifting effect is higher than 0% to the total number of users for the important action acquired in step S25. In step S29, the determination unit 29 determines that the important action with the highest ratio is effective. In step S29, the determination unit 29 determines that the important action is ineffective for important actions other than the important action with the highest ratio. Note that, although the ratio of users is calculated above, an average value of the lifting effect may also be calculated.

[0048] The response table generating unit 30 stores the important actions (excluding those without promotion effects) acquired in step S25 and the difference (praise effects) acquired in step S28 in the response table in association with the user ID.

[0049] In step S29, when an important action for the final action designated in step S22 is determined, a process for determining an important action for another final action is carried out.

[0050] FIG. 8 is a diagram showing the above process in a simplified manner. FIG. 8(a) is a schematic diagram showing learning data based on the behavior of all users. In the present disclosure, the learning data shows behavior over a certain period up to a specific date and time. For example, 500 pieces of learning data are prepared, each representing a user's behavior. FIG. 8(b) is learning data based on the behavior of user AAA. As with FIG. 8(a), there are 500 pieces of learning data representing behavior over a certain period up to a specific date.

[0051] FIG. 8(c) is a schematic diagram showing an outline of the process of generating the completion probability estimation model 26a and the answer table 103 based on these learning data.

[0052] As shown in FIG. 8(c), the model generation unit 26 inputs the learning data shown in FIG. 8(a) and generates a completion probability estimation model 26a for each final action. In the figure, the completion probability estimation models 26a for models 1 to x are generated. These models are, for example, models for estimating the probability of completing a model change or a probability of completing a plan change. Meanwhile, the important action acquisition unit 25 acquires important actions. Then, the answer table generation unit 30 generates an answer table 103 using the learning data, important actions, and completion probability estimation models 26a of user AAA.

[0053] Next, the output model 23 will be described. The output model 23 outputs the degree of association between one or more sub-actions acquired by the learning data acquisition unit 21 and a final action. The output model 23 outputs the degree of association between the sub-actions and the final action using known means. In the present disclosure, the output model 23 analyzes the sequence of actions up to the final action based on the user actions acquired by the learning data acquisition unit 21. As a method for the output model 23 to analyze the sequence of actions, any known method can be adopted as long as it is a method that can analyze the strength of association (contribution) between user actions recorded in a sequence.

[0054] For the above analysis, it is possible to apply a technology used in natural language processing, such as BERT. In this case, in natural language processing, the strength of association between tokens defined by words or the like is obtained as attention (degree of association). In this case, each user action is defined as a token, and an output model 23 may be generated by learning a model that outputs the magnitude of the degree of association between user actions.

[0055] 9 is a diagram showing an example of a process for calculating the degree of association between a sub-action and a final action. The output model 23 outputs the degree of association between the sub-action and the final action for each sub-action performed before the final action.

[0056] FIG. 9 shows an example in which a user performs sub-actions A, B, and C before the final action, "device change" procedure. The output model 23 outputs, for example, a value of 0.2 as the degree of association between sub-action A and the final action. The output model 23 outputs, for example, a value of 0.5 as the degree of association between sub-action B and the final action. The output model 23 outputs, for example, a value of 0.1 as the degree of association between sub-action C and the final action.

[0057] Then, the extraction unit 24 extracts a predetermined number of small actions with the highest degree of association from among the small actions performed by the users of the first user group.

[0058] Next, prediction of the timing of important actions will be explained. In this disclosure, large-scale action models (LAM) are used. LAM is based on Transformer, a neural network of sequence transformation models, and is designed to learn the relationships between actions arranged in a time series. Transformer is a technology that is increasingly being used in the field of natural language processing, and is a model that self-learns the relationships between tokens such as words. In contrast, LAM applies the mechanism of Transformer to the analysis of user actions.

[0059] FIG. 10 is a diagram showing the learning process of a large-scale behavioral model. As shown in the figure, the parameters of the large-scale behavioral model are trained using a masked language model. This involves replacing part of a sequence with a token representing a gap, such as [MASK], and then estimating what will go in that place based on surrounding tokens (behaviors), thereby learning the relationships between tokens. In the present disclosure, in order to improve the prediction accuracy of user behavior (tokens), in addition to random replacement, the rear of the sequence may also be replaced at the same time. This makes it possible to efficiently learn the relationships between subsequent tokens from the order of tokens at the front of the sequence. In this learning, each user behavior is replaced with a token in order, and the relationships between tokens in all patterns are learned.

[0060] When a sequence containing [MASK] is input into LAM, the probability of an action occurring at the time of [MASK] is predicted. By replacing [MASK] for each of the user's actions, it is possible to determine the specific action corresponding to that [MASK], its probability of occurrence, and its location (timing).

[0061] In the present disclosure, a position where the probability of an important behavior occurring is high is identified, and when an important behavior is inserted at that position, the completion probability is estimated, thereby making it possible to understand the effect.

[0062] The major difference between the two is that while the position of a word in a language is expressed as a relative position within the entire document, customer behavior is expressed as an absolute position based on date and time. In customer behavior analysis, this date and time information is important, so it is preferable to take date and time information into consideration when learning.

[0063] FIG. 11 is a diagram illustrating an example of a process for calculating the difference in completion probability. The completion probability acquisition unit 27 acquires, for each user, the completion probability of reaching a final action of a user action that does not include a significant action, using the completion probability estimation model 26a generated by the model generation unit 26. The completion probability acquisition unit 27 inputs the user action into the completion probability estimation model as a user action that does not include a significant action. As a result, the completion probability acquisition unit 27 acquires the completion probability of the user action that does not include a significant action. In FIG. 11, the significant action is inserted at a position where the probability of the significant action occurring is high, as described above, and the completion probabilities are compared.

[0064] Next, the effects of the dialogue device 100 of the present disclosure will be described. The dialogue device 100 of the present disclosure includes a response table 103 (storage unit) that stores response information including procedure details (final actions) and recommended items (additional actions) related to the procedure details. The dialogue unit 101 functions as an acquisition unit that acquires user inquiries and as an answering unit, and when an inquiry is received from the user terminal 200, it refers to the response table 103 and answers the procedure details (final actions) corresponding to the inquiry and the recommended items (additional actions) corresponding to the procedure details.

[0065] This configuration allows the user to be presented with recommended items related to the procedure they have inquired about, thereby increasing the probability that the user will complete the procedure. In other words, it is possible to respond to users whose requests are unclear, thereby improving user satisfaction.

[0066] In the present disclosure, the response table 103 further stores the effects of the recommended items. Then, the inquiry content acquisition unit 102 selects recommended items based on the effects, and the dialogue unit 101 transmits the recommended items to the user terminal 200.

[0067] For example, if the answer table 103 stores a plurality of recommended items and effects for the procedure content, the inquiry content acquisition unit 102 selects the recommended item with the highest effect, and the dialogue unit 101 answers with that recommended item.

[0068] According to this configuration, highly effective recommended items can be presented to the user, thereby increasing the probability that the user will complete the procedure.

[0069] In the present disclosure, the recommendation items are determined based on the completion probability of the procedure content (final action) corresponding to the inquiry, calculated based on the user's behavior history. For example, the behavior history is the user's behavior when visiting a store, web browsing, etc. In other words, the recommendation items indicate the procedures to complete the procedure content, thereby improving the user's ability to complete the procedure.

[0070] The recommendation items of the present disclosure are important actions that are highly relevant to the procedure content set in advance. For example, when a request is received for a procedure to change a rate plan, a procedure (action) for performing a rate simulation is an important action that is highly relevant.

[0071] In the present disclosure, the effectiveness of a recommended item is determined based on the completion probability when the recommended item is included in a predetermined procedure (final action) and the completion probability when the recommended item is not included in the procedure. These completion probabilities are calculated using a completion probability estimation model 26a.

[0072] In the present disclosure, the answer table 103 may be created in advance by the designer or operator of the dialogue device 100, or may be mechanically generated by an answer table generation device based on the user's dialogue history (user behavior including inquiries and their responses), web browsing, and behavior when visiting a store.

[0073] The answer table generation device 300 of the present disclosure is a generation device that generates answer information to be stored in the answer table 103 included in the dialogue device 100. The learning data acquisition unit 21 acquires user actions including multiple user actions and a final action such as the user's procedure content. Then, the output model 23 classifies the user using the classification unit 22, and based on the classification, outputs the degree of association between each of the user's multiple actions and the final action.

[0074] The important action acquisition unit 25 acquires an important action from one or more actions based on the degree of association. For example, an action with a high degree of association is regarded as an important action with respect to the final action.

[0075] The answer table generating unit 30 associates the acquired important action with the final action in the answer information and registers it in the answer table 103.

[0076] In this way, by setting actions that are highly related to the final action as important actions and setting them as additional actions, it is possible to present to the user additional actions that make it easier to mechanically achieve the final action.

[0077] In addition, the answer table generation device 300 of the present disclosure may have the following configuration. That is, the answer table generation device 300 includes a model generation unit 26 that generates a completion probability estimation model by machine learning using user actions as explanatory variables and reaching a final action as a target variable. The generated completion probability estimation model 26a acquires, for each user, the completion probabilities of reaching a final action for each of user actions including a significant action and user actions not including a significant action.

[0078] The calculation unit 28 calculates the difference between the completion probability of the user behavior including the important behavior and the completion probability of the user behavior not including the important behavior, and the judgment unit 29 judges whether the important behavior is effective based on the difference. The answer table generation unit 30 registers the important behavior in the answer table 103 if it is effective.

[0079] This allows effective additional actions to be registered in the response table 103, and additional actions utilizing this can be presented to the user.

[0080] The device and generating device of the present disclosure have the following configuration.

[0081] [1] a storage unit that stores answer information including a final action and an additional action related to the final action; an acquisition unit that acquires a user's inquiry; a response unit that, upon receiving the inquiry, refers to the storage unit and responds with a final action corresponding to the inquiry and an additional action corresponding to the final action; An apparatus comprising:

[0082] [2] the storage unit further stores the effect of the additional action; the response unit responds with the additional action based on the effect. [1] The device described in [1].

[0083] [3] When the memory unit stores a plurality of additional actions and effects for the final action, the answering unit answers the additional action with the greater effect. [2] The device described in [2].

[0084] [4] The additional action is determined based on a completion probability of a final action corresponding to the inquiry, the completion probability being calculated based on the user's action history. [1] to [3].

[0085] [5] The additional behavior is an important behavior that is highly related to a predetermined final behavior. [4] The device described in [4].

[0086] [6] The completion probability is calculated based on an estimation model trained using user actions of a plurality of users as explanatory variables and whether or not the plurality of users have reached a final action as a target variable. [4] The device described in [4].

[0087] [7] The effect of the additional action is determined based on a completion probability when the additional action is included in a preset final action and a completion probability when the additional action is not included in the preset final action. [2] The device according to [8] A generating device that generates answer information to be stored in the storage unit included in the device according to any one of [1] to [6], a learning data acquisition unit that acquires user actions including a plurality of user actions and a final action of the user; an output model that outputs a degree of association between each of the plurality of actions and the final action; an important behavior acquisition unit that acquires an important behavior from the plurality of behaviors based on the relevance; a generation unit that associates the important action with the final action in the response information as an additional action; A generating device comprising:

[0088] [9] a model generation unit that generates a completion probability estimation model by machine learning using the user's behavior as an explanatory variable and the user's reaching the final behavior as an objective variable; a completion probability acquisition unit that acquires, for each user, a completion probability of the user actions including the important action and the user actions not including the important action reaching the final action using the completion probability estimation model; a calculation unit that calculates a difference between the completion probability of the user behavior including the important behavior and the completion probability of the user behavior not including the important behavior; a determination unit that determines whether or not the important action is effective based on the difference; Further provided with [8] The generating device according to [8].

[0089]

[10] an acquisition step of acquiring a user query; a reply step of replying, upon receiving the inquiry, a final action and an additional action corresponding to the final action by referring to a storage unit that stores action information including a final action and an additional action related to the final action; A method comprising:

[0090] The block diagrams used to explain the above embodiments show functional blocks. These functional blocks (components) are realized by any combination of hardware and / or software. Furthermore, the method for realizing each functional block is not particularly limited. That is, each functional block may be realized using a single device that is physically or logically coupled, or may be realized using two or more physically or logically separated devices that are connected directly or indirectly (for example, by wire, wirelessly, etc.) and these multiple devices. The functional block may also be realized by combining the single device or multiple devices with software.

[0091] Functions include, but are not limited to, judgment, determination, discrimination, calculation, computation, processing, derivation, investigation, search, confirmation, reception, transmission, output, access, resolution, selection, selection, establishment, comparison, assumption, expectation, consideration, broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating, mapping, and assignment. For example, a functional block (component) that performs transmission is called a transmitting unit or transmitter. As mentioned above, there are no particular limitations on how these functions are implemented.

[0092] 12 is a diagram showing an example of the hardware configuration of the dialogue device 100 and the answer table generating device 300 according to this embodiment. The dialogue device 100 and the answer table generating device 300 described above may be physically configured as a computer device including a processor 1001, a memory 1002, a storage 1003, a communication device 1004, an input device 1005, an output device 1006, a bus 1007, etc.

[0093] In the following description, the word "apparatus" can be read as a circuit, a device, a unit, etc. The hardware configuration of the dialogue apparatus 100 and the answer table generating apparatus 300 may be configured to include one or more of the apparatuses shown in the drawings, or may be configured to exclude some of the apparatuses.

[0094] Each function of the dialogue device 100 and the response table generating device 300 is realized by loading predetermined software (programs) onto hardware such as the processor 1001 and memory 1002, causing the processor 1001 to perform calculations, control communication via the communication device 1004, and control at least one of reading and writing data from and to the memory 1002 and storage 1003.

[0095] The processor 1001 controls the entire computer by running, for example, an operating system. The processor 1001 may be configured by a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic unit, a register, etc. For example, the above-mentioned learning data acquisition unit 21 and the like may be realized by the processor 1001.

[0096] The processor 1001 also reads programs (program codes), software modules, data, etc. from at least one of the storage 1003 and the communication device 1004 into the memory 1002 and executes various processes in accordance with these. The program used is a program that causes a computer to execute at least some of the operations described in the above-described embodiments. For example, at least one of the functional units of the dialogue device 100 and the answer table generating device 300 may be implemented by a control program stored in the memory 1002 and running on the processor 1001, and similar implementations may be made for other functional blocks. While the above-described various processes have been described as being executed by one processor 1001, they may also be executed simultaneously or sequentially by two or more processors 1001. The processor 1001 may be implemented by one or more chips. The programs may also be transmitted from a network via a telecommunications line.

[0097] The memory 1002 is a computer-readable recording medium and may be configured, for example, by at least one of a read-only memory (ROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a random-access memory (RAM), etc. The memory 1002 may also be called a register, a cache, a main memory (primary storage device), etc. The memory 1002 can store executable programs (program codes), software modules, etc. for implementing the accompanying determination method according to one embodiment of the present disclosure.

[0098] Storage 1003 is a computer-readable recording medium, and may be, for example, at least one of an optical disk such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, a magneto-optical disk (e.g., a compact disk, a digital versatile disk, a Blu-ray disc), a smart card, a flash memory (e.g., a card, a stick, a key drive), a floppy disk, a magnetic strip, etc. Storage 1003 may also be referred to as an auxiliary storage device. The above-mentioned storage medium may be, for example, a database, a server, or other appropriate medium including at least one of memory 1002 and storage 1003.

[0099] The communication device 1004 is hardware (transmission / reception device) for communicating between computers via at least one of a wired network and a wireless network, and is also referred to as, for example, a network device, a network controller, a network card, or a communication module. The communication device 1004 may be configured to include a high-frequency switch, a duplexer, a filter, a frequency synthesizer, etc. to realize at least one of frequency division duplex (FDD) and time division duplex (TDD). For example, the above-mentioned dialogue unit 101 and the like may be realized by the communication device 1004. The communication device 1004 may be implemented with a transmitter and a receiver that are physically or logically separated from each other.

[0100] The input device 1005 is an input device (for example, a keyboard, a mouse, a microphone, a switch, a button, a sensor, etc.) that receives input from the outside. The output device 1006 is an output device (for example, a display, a speaker, an LED lamp, etc.) that outputs to the outside. The input device 1005 and the output device 1006 may be integrated into one device (for example, a touch panel).

[0101] Furthermore, each device, such as the processor 1001 and the memory 1002, is connected by a bus 1007 for communicating information. The bus 1007 may be configured using a single bus, or may be configured using different buses between each device.

[0102] Furthermore, the dialogue device 100 and the response table generating device 300 may be configured to include hardware such as a microprocessor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a field-programmable gate array (FPGA), and some or all of the functional blocks may be realized by the hardware. For example, the processor 1001 may be implemented using at least one of these pieces of hardware.

[0103] The notification of information is not limited to the aspects / embodiments described in the present disclosure and may be performed using other methods. For example, the notification of information may be performed by physical layer signaling (e.g., Downlink Control Information (DCI), Uplink Control Information (UCI)), higher layer signaling (e.g., Radio Resource Control (RRC) signaling, Medium Access Control (MAC) signaling, broadcast information (Master Information Block (MIB), System Information Block (SIB))), other signals, or a combination thereof. Furthermore, the RRC signaling may be referred to as an RRC message, and may be, for example, an RRC Connection Setup message, an RRC Connection Reconfiguration message, or the like.

[0104] The order of the procedures, sequences, flowcharts, etc. of each aspect / embodiment described in this disclosure may be changed unless it is consistent. For example, the methods described in this disclosure present elements of various steps using an example order, and are not limited to the particular order presented.

[0105] Input and output information may be stored in a specific location (for example, memory) or may be managed using a management table. Input and output information may be overwritten, updated, or added to. Output information may be deleted. Input information may be sent to another device.

[0106] The determination may be made based on a value represented by one bit (0 or 1), a Boolean value (true or false), or a numerical comparison (e.g., comparison with a predetermined value).

[0107] Each aspect / embodiment described in this disclosure may be used alone, in combination, or switched depending on the implementation. Furthermore, notification of predetermined information (e.g., notification that "X is true") is not limited to being done explicitly, but may be done implicitly (e.g., by not notifying the predetermined information).

[0108] Although the present disclosure has been described in detail above, it is clear to those skilled in the art that the present disclosure is not limited to the embodiments described herein. The present disclosure can be implemented in modified and altered forms without departing from the spirit and scope of the present disclosure as defined by the claims. Therefore, the description of the present disclosure is intended to be illustrative and does not have any limiting meaning on the present disclosure.

[0109] Software shall be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, threads of execution, procedures, functions, etc., whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise.

[0110] Software, instructions, information, etc. may also be transmitted or received over a transmission medium. For example, if software is transmitted from a website, server, or other remote source using wired technologies (such as coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL)), and / or wireless technologies (such as infrared, microwave), these wired and / or wireless technologies are included within the definition of transmission media.

[0111] The information, signals, etc. described in this disclosure may be represented using any of a variety of different technologies. For example, data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may be represented by voltages, currents, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.

[0112] Note that terms explained in this disclosure and terms necessary for understanding this disclosure may be replaced with terms having the same or similar meanings. For example, at least one of a channel and a symbol may be a signal (signaling). Furthermore, a signal may be a message. Furthermore, a component carrier (CC) may be called a carrier frequency, a cell, a frequency carrier, etc.

[0113] Furthermore, the information, parameters, etc. described in the present disclosure may be expressed using absolute values, may be expressed using relative values ​​from a predetermined value, or may be expressed using other corresponding information. For example, a radio resource may be indicated by an index.

[0114] The names used for the above-described parameters are not intended to be limiting in any way. Furthermore, the mathematical expressions using these parameters may differ from those explicitly disclosed in this disclosure. The various channels (e.g., PUCCH, PDCCH, etc.) and information elements may be identified by any suitable names, and therefore the various names assigned to these various channels and information elements are not intended to be limiting in any way.

[0115] In this disclosure, the terms "Mobile Station (MS)," "user terminal," "User Equipment (UE)," "terminal," etc. may be used interchangeably.

[0116] A mobile station may also be referred to by those skilled in the art as a subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, or some other suitable terminology.

[0117] As used in this disclosure, the terms "determining" and "determining" may encompass a wide variety of actions. "Determining" and "determining" may include, for example, judging, calculating, computing, processing, deriving, investigating, looking up, searching, inquiring (e.g., searching in a table, database, or other data structure), ascertaining, and the like. "Determining" and "determining" may also include receiving (e.g., receiving information), transmitting (e.g., sending information), input, output, accessing (e.g., accessing data in memory), and the like. Furthermore, "judgment" and "decision" can include regarding resolving, selecting, choosing, establishing, comparing, etc. as having been "judged" or "decided." In other words, "judgment" and "decision" can include regarding some action as having been "judged" or "decided." Furthermore, "judgment (decision)" can be interpreted as "assuming," "expecting," "considering," etc.

[0118] The terms "connected," "coupled," or any variation thereof, refer to any direct or indirect connection or coupling between two or more elements, and may include the presence of one or more intermediate elements between two elements that are "connected" or "coupled" to each other. The coupling or connection between elements may be physical, logical, or a combination thereof. For example, "connected" may be read as "access." As used in this disclosure, two elements may be considered to be "connected" or "coupled" to each other using one or more wires, cables, and / or printed electrical connections, as well as electromagnetic energy having wavelengths in the radio frequency range, microwave range, and optical (both visible and invisible) range, as some non-limiting and non-exhaustive examples.

[0119] As used in this disclosure, the phrase "based on" does not mean "based only on," unless expressly stated otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

[0120] Any reference to an element using a designation such as "first," "second," etc., used in this disclosure does not generally limit the quantity or order of those elements. These designations may be used in this disclosure as a convenient way to distinguish between two or more elements. Thus, a reference to a first and a second element does not imply that only two elements may be employed or that the first element must in some way precede the second element.

[0121] When used in this disclosure, the terms "include," "including," and variations thereof are intended to be inclusive, similar to the term "comprising." Furthermore, when used in this disclosure, the term "or" is not intended to be an exclusive or.

[0122] In this disclosure, where articles are added by translation, such as a, an, and the in English, the disclosure may include that the nouns following these articles are in the plural form.

[0123] In the present disclosure, the term "A and B are different" may mean "A and B are different from each other." The term may also mean "A and B are each different from C." Terms such as "separate" and "coupled" may also be interpreted in the same way as "different." [Explanation of symbols]

[0124] 100...Dialogue device, 101...Dialogue unit, 102...Query content acquisition unit, 103...Answer table, 200...User terminal, 300...Answer table generation device, 21...Learning data acquisition unit, 22...Classification unit, 23...Output model, 24...Extraction unit, 25...Important behavior acquisition unit, 26...Model generation unit, 27...Completion probability acquisition unit, 28...Calculation unit, 29...Judgment unit.

Claims

1. a storage unit that stores answer information including a final action and an additional action related to the final action; an acquisition unit that acquires a user's inquiry; a response unit that, upon receiving the inquiry, refers to the storage unit and responds with a final action corresponding to the inquiry and an additional action corresponding to the final action; An apparatus comprising:

2. the storage unit further stores an effect of the additional action; the response unit responds with the additional action based on the effect.

10. The apparatus of claim 1.

3. When the memory unit stores a plurality of additional actions and effects for the final action, the answering unit answers the additional action with the greater effect.

3. The apparatus of claim 2.

4. The additional action is determined based on a completion probability of a final action corresponding to the inquiry, the completion probability being calculated based on the user's action history.

10. The apparatus of claim 1.

5. The additional behavior is an important behavior that is highly related to a predetermined final behavior.

5. The apparatus of claim 4.

6. The completion probability is calculated based on an estimation model trained using user actions of a plurality of users as explanatory variables and whether or not the plurality of users have reached a final action as a target variable.

5. The apparatus of claim 4.

7. The effect of the additional action is determined based on a completion probability when the additional action is included in a preset final action and a completion probability when the additional action is not included in the preset final action.

3. The apparatus of claim 2.

8. A generating device that generates answer information to be stored in the storage unit included in the device according to claim 1, a learning data acquisition unit that acquires user actions including a plurality of user actions and a final action of the user; an output model that outputs a degree of association between each of the plurality of actions and the final action; an important behavior acquisition unit that acquires an important behavior from the plurality of behaviors based on the relevance; a generation unit that associates the important action with the final action in the response information; A generating device comprising:

9. a model generation unit that generates a completion probability estimation model by machine learning using the user's behavior as an explanatory variable and the user's reaching the final behavior as an objective variable; a completion probability acquisition unit that acquires, for each user, a completion probability of the user actions including the important action and the user actions not including the important action reaching the final action using the completion probability estimation model; a calculation unit that calculates a difference between the completion probability of the user behavior including the important behavior and the completion probability of the user behavior not including the important behavior; a determination unit that determines whether or not the important action is effective based on the difference; The generating device of claim 8 further comprising:

10. an acquisition step of acquiring a user query; a reply step of replying, upon receiving the inquiry, a final action and an additional action corresponding to the final action by referring to a storage unit that stores action information including a final action and an additional action related to the final action; A method comprising:

Images