Method for recommendation

Abstract

To improve a rule related to recommendation.SOLUTION: A method for recommendation is a method of recommending a stop-off point to a user based on a predetermined rule. The method for recommendation includes: a reception step of receiving input of feedback information from a user, showing whether the recommendation of a stop-off point is appropriate; a determination step of determining whether the feedback information satisfies an improvement condition related to a predetermined rule; and an update step of updating the predetermined rule on the basis of the feedback information in a case where the feedback information has been determined to satisfy the improvement condition.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to the technical field of a recommendation method for recommending stop-off points. [Background technology]

[0002] As an example of this type of method, a method has been proposed in which a user's mental stress based on the user's movement history is used as a search condition to determine a suggested destination (see Patent Document 1). Other prior art documents related to the present invention include Patent Documents 2 to 5. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2014-098623 A [Patent Document 2] JP 2020-052468 A [Patent Document 3] JP 2008-233033 A [Patent Document 4] JP 2019-191683 A [Patent Document 5] JP 2020-153919 A Summary of the Invention [Problem to be solved by the invention]

[0004] For example, the technology described in Patent Document 1 does not have a configuration for utilizing feedback from a user regarding a proposed destination. Therefore, the technology described in Patent Document 1 has a technical problem in that even if a user judges that a proposed destination as a proposed result is not appropriate, the proposed result cannot be improved.

[0005] The present invention has been made in consideration of the above problems, and an object of the present invention is to provide a recommendation method that can improve rules relating to recommendation. [Means for solving the problem]

[0006] A recommendation method according to one aspect of the present invention is a recommendation method for recommending stop-off spots to a user based on predetermined rules, and includes a receiving step for receiving input of feedback information from the user indicating whether or not the recommendation of the stop-off spot was appropriate, a determination step for determining whether or not the feedback information satisfies an improvement condition related to the predetermined rule, and an update step for updating the predetermined rule based on the feedback information when it is determined that the feedback information satisfies the improvement condition. [Brief description of the drawings]

[0007] [Figure 1] 1 is a block diagram showing a configuration of a vehicle according to an embodiment; [Diagram 2] 4 is a flowchart showing an operation of the navigation device according to the embodiment. [Diagram 3] FIG. 13 is a diagram showing an example of a display image. [Figure 4] FIG. 13 is a diagram showing another example of a display image. [Diagram 5] FIG. 13 is a diagram showing another example of a display image. [Figure 6] 10 is a flowchart illustrating a feedback operation according to the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] An embodiment of the recommendation method will be described with reference to FIGS.

[0009] (Vehicle configuration) The configuration of a vehicle according to the embodiment will be described with reference to Fig. 1. In Fig. 1, a vehicle 1 includes an exterior communication device 11, a position detection device 12, an on-board camera 13, an on-board sensor 14, and a navigation device 20. The vehicle 1 may be a vehicle having an automatic driving function.

[0010] The external communication device 11 is a device having a wireless communication function. The external communication device 11 is connected to a network via a wireless base station. The position detection device 12 is a device that detects the position of the vehicle 1. The position detection device 12 may be, for example, a GPS (Global Positioning System) receiver.

[0011] The on-board camera 13 is not limited to a camera that captures the surroundings of the vehicle 1 (in other words, the outside of the vehicle 1), but may be a camera that captures the cabin of the vehicle 1 (in other words, the interior of the vehicle 1). The on-board camera 13 may include two or more cameras. In this case, the on-board camera 13 may include a camera that captures the surroundings of the vehicle 1 and a camera that captures the cabin of the vehicle 1. The on-board sensor 14 may include, for example, at least one of a speed sensor, an acceleration sensor, a yaw rate sensor, a steering angle sensor, a temperature sensor, a humidity sensor, and a rainfall sensor.

[0012] The navigation device 20 includes a control unit 21, a storage unit 22, an in-vehicle communication interface (I / F) 23, a display unit 24, and an operation unit 25. The control unit 21 may include one or more processors that execute a predetermined computer program. The storage unit 22 may be a storage medium that stores the predetermined computer program executed by the control unit 21. The in-vehicle communication I / F 23 may be a communication interface circuit that enables the navigation device 20 to communicate with other in-vehicle devices of the vehicle 1 via an in-vehicle network.

[0013] The display unit 24 may be, for example, a liquid crystal display or an organic EL (Electro-Luminesence) display. The operation unit 25 may be, for example, at least one of a touch panel, a mouse, a keyboard, and an operation button. The display unit 24 and the operation unit 25 may be integrated as a touch display.

[0014] (Operation of navigation device) The operation of the navigation device 20 will be described with reference to the flowchart in Fig. 2. First, an overview of the operation of the navigation device 20 will be described. After that, the operation of the navigation device 20 will be described with specific examples.

[0015] In FIG. 2, the control unit 21 of the navigation device 20 selects a feature for recommending a stop-off spot and calculates a weighting parameter for the selected feature (step S101). The feature may include, for example, at least one of the current time, season, weather, vehicle data, road information, business hours, site area, parking information, type of stop-off spot, and popularity. The popularity may be calculated based on at least one of the number of visitors and the number of visits by each person. The control unit 21 may calculate the weighting parameter by applying a predetermined filtering process to the category feature. The control unit 21 may calculate the weighting parameter by applying, for example, Laplace smoothing to the number feature.

[0016] Next, the control unit 21 acquires the selection criteria (step S102). For example, the control unit 21 may change the selection criteria based on at least one of the type of stop-off place to be recommended and the feature amount selected in the processing of step S101. The selection criteria may include a calculation formula for merging the feature amount selected in the processing of step S101. The selection criteria may include the number of stop-off places to be recommended to the user of the vehicle 1. The selection criteria may include an exclusion filter for determining stop-off places to be excluded. The selection criteria may be a learning model that outputs stop-off places to be recommended to the user when the feature amount is input. The learning model may be called a recommendation model because it is used to recommend stop-off places.

[0017] Next, the control unit 21 generates a recommendation result and a recommendation reason (step S103). For example, the control unit 21 may calculate a score for each of the multiple stop-off location candidates using a calculation formula included in the selection criteria. The control unit 21 may select one or more stop-off location candidates to be recommended to the user based on the calculated score. The control unit 21 may generate information indicating the selected one or more stop-off locations as the recommendation result. For example, the control unit 21 may extract one or more feature amounts that have a large influence on the score from among multiple feature amounts related to the selected one or more stop-off locations. The control unit 21 may generate the recommendation reason based on the extracted one or more feature amounts.

[0018] The control unit 21 determines the timing for displaying the recommendation result and the recommendation reason (step S104). The process of step S104 may be performed in parallel with the processes of steps S101 to S103. For example, the control unit 21 may determine the timing for displaying the recommendation result and the recommendation reason as being a timing that does not affect the driving of the vehicle 1. The timing that does not affect the driving of the vehicle 1 may include at least one of the time when the shift lever is in the "P" position and the time when the vehicle 1 is stopped.

[0019] The control unit 21 controls the display unit 24 to display the recommendation result and the reason for recommendation at the display timing determined in the processing of step S104 (step S105). The control unit 21 also controls the display unit 24 to display the evaluation content. The "evaluation content" may mean the user's evaluation items for the recommendation result. Specific examples of the "recommendation result and reason for recommendation" and the "evaluation content" will be described later. The display timing of the "recommendation result and reason for recommendation" and the display timing of the "evaluation content" may be different or the same.

[0020] The control unit 21 determines whether there is feedback from the user (step S106). If it is determined in the process of step S106 that there is no feedback from the user (step S106: Yes), the operation shown in Fig. 2 is ended. The process of step S101 may be performed after a first predetermined time has elapsed since the operation shown in Fig. 2 ended. In other words, the operation shown in Fig. 2 may be repeatedly performed at a period corresponding to the first predetermined time.

[0021] In the process of step S106, if it is determined that there is user feedback (step S106: Yes), the control unit 21 improves the recommendation logic related to the stop-off points based on the feedback (step S107). A specific example of a method for improving the recommendation logic will be described later. Improving the recommendation logic may mean improving at least a part of the above-mentioned selection criteria.

[0022] (First concrete example) As a first specific example of the operation of the navigation device 20, an operation of recommending a stopover place for spending leisure time will be described with reference to Fig. 2 as well as Fig. 3 to Fig. 6. The stopover place for spending leisure time may include at least one of a restaurant and a tourist spot.

[0023] For example, the control unit 21 may acquire visit history data of at least one of restaurants and tourist spots of the user of the vehicle 1. For example, the control unit 21 may acquire visit history data of at least one of restaurants and tourist spots of another person (i.e., a person other than the user of the vehicle 1).

[0024] In the process of step S101 in FIG. 2, the control unit 21 may select, as a feature amount, at least one of the genre of a restaurant visited by the user (e.g., Japanese food, Western food, Chinese food, etc.) and the genre of a tourist spot visited by the user (e.g., nature, leisure facilities, historical buildings, etc.) based on the visit history data of the user of the vehicle 1. The control unit 21 may select, as a feature amount, a transition between places indicating that the user visited one place and then visited another place based on the visit history data of the user of the vehicle 1. The control unit 21 may select, as a feature amount, at least one of the number of people who visited one restaurant and the number of people who visited one tourist spot based on the visit history data of another person. The control unit 21 may select, as a feature amount, a transition between places indicating that the other person visited one place and then visited another place based on the visit history data of the other person. The control unit 21 may select, as a feature amount, at least one of the current time, season, weather, vehicle data, road information, business hours, site area, and parking lot information.

[0025] For example, if the visit history data of the user of vehicle 1 does not include a situation that matches or is similar to the current situation of vehicle 1 (e.g., driving location, time of day, weather, etc.), the control unit 21 does not need to use the visit history data of the user of vehicle 1 in the processing of step S101.

[0026] In the process of step S102 in FIG. 2, the control unit 21 may acquire, for example, a formula such as "recommendation score=x1×α+x2×β+···" as the selection criterion. In the above formula, "x1, x2,···" represent features, and "α, β,···" represent parameters (e.g., weighting parameters). The above formula corresponds to an example of a calculation formula for merging the features selected in the process of step S101. In the process of step S102, the control unit 21 may acquire, for example, an exclusion filter such as "exclude outdoor facilities when it is raining" or "exclude facilities that are closed for business hours" as the selection criterion.

[0027] In the process of step S105 in FIG. 2, the control unit 21 controls the display unit 24 so that the recommendation result and the recommendation reason are displayed. In this case, the display unit 24 may display an image as shown in FIG. 3, for example. In FIG. 3, a triangle C indicates the vehicle 1, and a solid line R indicates the road on which the vehicle 1 is traveling. The "recommended place" shown in the upper right of FIG. 3 is an example of the recommendation result and the recommendation reason. In FIG. 3, the numbers enclosed in squares around the solid line R correspond to the positions of the recommended places "Top 1", "Top 2", and "Top 3", respectively. The "popularity", "distance", "preference", and "set visit" included in the recommended place correspond to an example of the recommendation reason. As shown in FIG. 3, the control unit 21 may control the display unit 24 so that the recommendation result and the recommendation reason are displayed on an image for performing route guidance. The display method of the recommendation result and the recommendation reason as shown in FIG. 3 may be called a push-type notification method.

[0028] Alternatively, in the process of step S105, the display unit 24 may display an image as shown in FIG. 4, for example. For example, when the operation shown in FIG. 2 is performed due to the user of the vehicle 1 inputting "recommendations around Mt. Fuji" using the operation unit 25, the display unit 24 may display an image as shown in FIG. 4. The display method of the recommendation result and the recommendation reason as shown in FIG. 4 may be called a pull-type notification method. For example, when the user of the vehicle 1 performs a place search using the operation unit 25, the control unit 21 may control the display unit 24 to display the recommendation result and the recommendation reason using the pull-type notification method. For example, when the user of the vehicle 1 does not set a destination in the navigation device 20, the control unit 21 may control the display unit 24 to display the recommendation result and the recommendation reason using the push-type notification method.

[0029] In the process of step S105, after the recommendation result and the reason for recommendation are displayed, the control unit 21 may control the display unit 24 to display the evaluation content. For example, when the user of the vehicle 1 selects "Top1: △△ Store" shown in FIG. 3 or FIG. 4, the display unit 24 may display an image for a questionnaire survey as shown in FIG. 5. Note that "1. Are the timing and location of the current recommendation information appropriate?", "2. Are you satisfied with the current recommendation result?", and "3. Which element did you mainly consider?" in FIG. 5 correspond to examples of the evaluation content. Note that the control unit 21 may control the display unit 24 to display the evaluation content even if the user of the vehicle 1 does not operate the operation unit 25 until a second predetermined time has elapsed since the recommendation result and the reason for recommendation were displayed (in other words, the user does not react to the recommendation result).

[0030] For example, if the user of the vehicle 1 responds to a questionnaire using the operation unit 25 after an image as shown in Fig. 5 is displayed, the control unit 21 may determine that there is user feedback in the process of step S106 in Fig. 2. On the other hand, if the user does not respond to the questionnaire, the control unit 21 may determine that there is no user feedback in the process of step S106.

[0031] An example of the process of step S107 in Fig. 2 will be described with reference to the flowchart in Fig. 6. Here, an example of the process of step S107 in the case where the selection criterion is a learning model will be described. In the process of step S107, the control unit 21 may improve the recommendation logic by performing reinforcement learning of the learning model using, for example, RLHF (Reinforcement Learning from Human Feedback).

[0032] 6, the control unit 21 processes feedback (e.g., responses to the questionnaire shown in FIG. 5) of the user of the vehicle 1 (step S201). For example, the control unit 21 may generate a policy function for adjusting parameters (e.g., weighting parameters) related to feature amounts based on the user feedback.

[0033] Next, the control unit 21 determines whether or not to improve the rule related to the recommendation logic (step S202). For example, the control unit 21 may obtain a change in the satisfaction level of the user of the vehicle 1 from the satisfaction level based on the previous feedback and the satisfaction level based on the current feedback. If the change in the satisfaction level is greater than a predetermined threshold, the control unit 21 may determine not to improve the rule. On the other hand, if the change in the satisfaction level is smaller than the predetermined threshold, the control unit 21 may determine to improve the rule. Note that, if there is no previous feedback, the change in the user's satisfaction level may be obtained from an initial value related to the satisfaction level and the satisfaction level based on the current feedback. Note that, since the rule is improved when the change in the satisfaction level is greater than the threshold, the change in the satisfaction level being greater than the threshold may be referred to as an improvement condition.

[0034] If it is determined in the processing of step S202 that the rule is not to be improved (step S202: No), for example, the control unit 21 stores the policy function generated in the processing of step S201 and terminates the processing shown in FIG. 6 (in other words, the processing of step S107 in FIG. 2 is terminated).

[0035] If it is determined in the process of step S202 that the rule is to be improved (step S202: Yes), the control unit 21 changes the rule based on the policy function generated in the process of step S201 (step S203). The control unit 21 may obtain a correspondence relationship between a change in a parameter related to the rule (e.g., a parameter related to a feature amount) and a change in the user's satisfaction level. The control unit 21 may change the rule by estimating a parameter that causes a change in the user's satisfaction level to exceed a predetermined threshold based on the obtained correspondence relationship.

[0036] A specific example of the process of step S203 will be described below. The learning model used to generate the recommendation result and the recommendation reason in the process of step S103 in Fig. 2 is set as the initial learning model. The control unit 21 may generate a tuned learning model based on the policy function and the initial learning model. For example, the control unit 21 may generate a tuned learning model using PPO (Proximal Policy Optimization) or TRPO (Trust Region Policy Optimization).

[0037] For example, the control unit 21 may input the feature quantity selected in the process of step S101 in FIG. 2 to the initial learning model and the tuned learning model. The control unit 21 may obtain a reward value by inputting the output result of the tuned learning model to a predetermined reward function. The control unit 21 may obtain a penalty corresponding to the difference between the output result of the initial learning model and the output result of the tuned learning model. The control unit 21 may correct the one reward value by using the penalty. The control unit 21 may change the policy function generated in the process of step S201 by using the corrected one reward value. The policy function may correspond to an example of a rule related to the recommendation logic. Improving the recommendation logic may mean changing the policy function.

[0038] After the process of step S203, the control unit 21 stores the changed rule (e.g., the changed policy function described above) in the storage unit 22 (step S204). In parallel with the process of step S204, the control unit 21 may control the display unit 24 so that the changed rule is presented to the user of the vehicle 1 (step S205). In this case, the control unit 21 changes the changed rule (e.g., the changed policy function described above) to a form that the user can understand. The form that the user can understand may include, for example, a graph showing the effect of stochastic gradient descent (e.g., how the value of the objective function converges). The form that the user can understand may include, for example, an explanatory text explaining the changed rule. The process of step S205 may be performed only when a request is made by the user using the operation unit 25, for example.

[0039] (Second concrete example) As a second specific example of the operation of the navigation device 20, an operation of recommending a stop-off point where a rest can be taken will be described with reference to Fig. 2. Note that, regarding the second specific example, a description that overlaps with the description of the first specific example described above will be omitted as appropriate.

[0040] For example, the control unit 21 may acquire driving history data of the user of the vehicle 1. The control unit 21 may acquire at least one of data indicating the behavior of the vehicle 1 and data indicating the state of the user of the vehicle 1. In the process of step S101 in FIG. 2, the control unit 21 may select the driving time of the user and the timing of stopping as the feature amount from the acquired driving history data. The control unit 21 may select information indicating the driving state of the user (e.g., the number of sudden braking, the number of sudden accelerations, the number of yawns, etc.) as the feature amount from at least one of data indicating the behavior of the vehicle 1 and data indicating the state of the user of the vehicle 1. The control unit 21 may select information related to the external environment of the vehicle 1 (e.g., road information, weather, etc.) as the feature amount. The control unit 21 may select attribute data related to a place where the user of the vehicle 1 can take a rest (e.g., the presence or absence of a toilet, the presence or absence of a parking lot, etc.) as the feature amount. The control unit 21 may select at least one of a destination set in the navigation device 20, a time required to reach the destination, and a position of the vehicle 1 at a certain time as the feature amount.

[0041] 2, the control unit 21 may acquire, for example, a formula such as "recommendation score=x1×α+x2×β+···" as the selection criterion. In the process of step S102, the control unit 21 may acquire, for example, an exclusion filter such as "excluding facilities that are not suitable for short breaks" or "excluding facilities that are not suitable for long breaks" as the selection criterion.

[0042] For example, the control unit 21 may estimate the timing of the user's rest based on at least one of driving history data of the user of the vehicle 1, a destination set in the navigation device 20, road information (e.g., traffic congestion information, construction information, etc.), and the location of a stopover point where a rest can be taken. The control unit 21 may include the estimated timing of the rest in the selection criterion.

[0043] The recommendation reasons generated in the process of step S103 in FIG. 2 may include, for example, at least one of "close distance," "restroom available," "dining available," "smoking available," "nap available," and "lodging available."

[0044] (Third Specific Example) As a third specific example of the operation of the navigation device 20, an operation of recommending a stop-off point where vehicle maintenance is possible will be described with reference to Fig. 2. Note that, for the third specific example, a description that overlaps with the description of the first specific example described above will be omitted as appropriate.

[0045] In the process of step S101 in Fig. 2, the control unit 21 may select data related to parts of the vehicle 1 as feature quantities. The data related to the parts is not limited to data directly indicating the state of the parts, and may be data indirectly indicating the state of the parts. For example, the state of the engine may be indicated as the temperature of the coolant that cools the engine. Note that the data related to consumable parts may be indicated by the mileage of the vehicle 1, or may be indicated by the time elapsed since the last replacement.

[0046] The control unit 21 may acquire at least one of the vehicle inspection history data, the car wash history data, and the maintenance history data of the vehicle 1. The control unit 21 may select the vehicle inspection date and time and the vehicle inspection location as the feature amount based on the vehicle inspection history data. The control unit 21 may select the car wash date and time and the car wash location as the feature amount based on the car wash history data. The control unit 21 may select the maintenance date and time and the maintenance location as the feature amount based on the maintenance history data.

[0047] 2, the control unit 21 may acquire, for example, a formula such as "recommendation score = x1 × α + x2 × β + ···" as the selection criterion. In the process of step S102, the control unit 21 may acquire, for example, an exclusion filter such as "exclude facilities that are not suitable for vehicle maintenance" as the selection criterion.

[0048] 2, the control unit 21 may perform abnormality detection for the parts of the vehicle 1. For example, the control unit 21 may perform abnormality detection for the parts of the vehicle 1 using a One Class Support Vector Machine. For example, the control unit 21 may classify the degree of abnormality, which indicates the degree of abnormality, into "high", "medium", or "low" according to the number of abnormality data.

[0049] The recommendation reason generated in the process of step S103 in Fig. 2 may include at least one of "close distance", "frequently used", and "highly rated". In the process of step S105 in Fig. 2, the control unit 21 may control the display unit 24 so that the display mode of the recommendation result changes depending on the abnormality degree. For example, when the abnormality degree is "high", the control unit 21 may control the display unit 24 so that a stop-off place where vehicle maintenance is possible is displayed together with a message "Please take emergency action". For example, when the abnormality degree is "medium", the control unit 21 may control the display unit 24 so that a stop-off place where vehicle maintenance is possible is displayed together with a message "Caution information".

[0050] (Technical effect) In the navigation device 20, when there is feedback from the user of the vehicle 1, the rules related to the recommendation logic are improved. Therefore, according to the method of recommending stop-off places by the navigation device 20, the rules related to the recommendation logic can be improved. As a result, it is possible to recommend more appropriate stop-off places.

[0051] Aspects of the invention derived from the above-described embodiment will be described below.

[0052] A recommendation method according to one aspect of the invention is a recommendation method for recommending stop-off spots to a user based on predetermined rules, and includes a receiving step for receiving input of feedback information from the user indicating whether or not the recommendation of the stop-off spot was appropriate, a determination step for determining whether or not the feedback information satisfies an improvement condition related to the predetermined rule, and an update step for updating the predetermined rule based on the feedback information when it is determined that the feedback information satisfies the improvement condition.

[0053] In the above-described embodiment, for example, displaying an image for a questionnaire survey as shown in FIG. 5 corresponds to an example of a reception process, the processing of step S202 in FIG. 6 corresponds to an example of a judgment process, and the processing of step S203 in FIG. 6 corresponds to an example of an update process.

[0054] The recommendation method may include a rule presenting step of presenting the rule updated in the update step to the user. In the above-described embodiment, the process of step S205 in FIG. 6 corresponds to an example of the presenting step.

[0055] The recommendation method may include a reason presenting step of presenting a reason for recommending the stop-off spot to the user. In the above-described embodiment, the process of step S105 in Fig. 2 corresponds to an example of the reason presenting step.

[0056] In the recommendation method, the user may be in a vehicle, and the recommended stop-off points may be presented to the user at a timing that does not affect the traveling of the vehicle.

[0057] The present invention is not limited to the above-described embodiments, but may be modified as appropriate within the scope of the claims and the gist or concept of the invention as can be read from the entire specification, and recommendation methods involving such modifications are also included in the technical scope of the present invention. [Explanation of symbols]

[0058] 1...vehicle, 20...navigation device, 21...control unit, 22...storage unit, 23...in-vehicle communication interface, 24...display unit, 25...operation unit

Claims

1. A recommendation method for recommending a stop-off point to a user based on a predetermined rule, comprising: a receiving step of receiving feedback information input by the user, the feedback information indicating whether the recommendation of the stop-off spot is appropriate or not; a determination step of determining whether the feedback information satisfies an improvement condition related to the predetermined rule; an updating step of updating the predetermined rule based on the feedback information when it is determined that the feedback information satisfies the improvement condition; Recommendation methods including.

2. a rule presenting step of presenting the rule updated in the update step to the user. The recommendation method according to claim 1.

3. A reason presentation step of presenting the reason for recommending the stop-off spot to the user. The recommendation method according to claim 1.

4. The user is in a vehicle, The recommended stopovers are presented to the user at a timing that does not affect the running of the vehicle. The recommendation method according to claim 1.

Images