JPWO2015155885A1 - Purchase history classification system and purchase history classification method - Google Patents

Abstract

A method of classifying a purchase history comprising a computer having a processor and a memory, and a terminal connected to the computer, wherein the computer identifies an identifier for identifying an individual and an identifier for a meal eaten or consumed by the individual. Receiving and storing meal information including, storing the meal feature quantity received from the terminal in the meal feature quantity storage unit, selecting the meal information for each individual identifier, and selecting the feature quantity corresponding to the meal information; Obtaining and analyzing the tendency of individual meal selection, calculating the meal selection probability as a food preference model based on the analysis result of the individual meal selection tendency for each individual, accepting input of meal information, When the number of people who eat and drink information is plural, the meal selected by the individual is estimated from the meal information based on the meal preference model.

Description

  The present invention relates to a purchase history classification system that analyzes meal purchase information and provides a method for accurately and simply recording information about meals actually eaten.

  In recent years, attention has been focused on improving dietary habits for the purpose of preventing lifestyle-related diseases such as metabolic syndrome, cerebrovascular disease, heart disease, diabetes and the like. To improve eating habits, it is important to record and visualize the contents of daily meals. However, recording daily meals itself is a very time-consuming task, and managing each meal from a nutritional point of view requires considerable effort, making it difficult for individuals to continue to work on. .

  Therefore, it is important to simplify meal management. For example, it is very effective if information on purchased products can be automatically collected and managed. In Patent Document 1, as a method for collecting information purchased by an individual, a method using personal identifier information included in a purchase history at the time of settlement is described.

US Pat. No. 8,452,653

  In the above-described conventional example, there is no problem when the user purchases only a product for himself, but the classification when the product is purchased including others is not considered. For example, when the product is assumed to be a meal, it cannot be said that the accountant and the eating (eating and drinking) person are the same. In the method of classifying products according to the individual identifier included in the purchase history at the time of payment in the prior art, the contents of meals that are different from actual ones may be recorded. That is, when one of the plurality of eaters settles the food and beverage charges of the plurality of eaters, in the above-described conventional example, all the eaters' food and beverages (products) are linked to the settled individual identifiers. There was a problem that.

  Therefore, the present invention determines whether the meal information included in the purchase history is related to the purchase history associated with one personal identifier, is consumed by the associated user, or includes information about other people, The purpose is to classify the information included in the purchase history when other people's information is also included.

  The present invention is a purchase history classification system comprising a computer including a processor and a memory, and a terminal connected to the computer via a network, the computer including an identifier for identifying an individual, the personal A meal information storage unit that receives and accumulates meal information including an identifier of a meal eaten or eaten from the terminal, a meal feature quantity storage unit that stores the meal feature quantity received from the terminal, and the individual identifier Each time the meal information is selected from the meal information storage unit, the feature amount corresponding to the meal information is acquired from the meal feature amount storage unit, and the tendency of the individual meal selection is analyzed. A food preference model construction unit that calculates the meal selection probability as a food preference model based on the analysis result of the meal selection tendency for each individual, and accepts input of the meal information An input unit, a determination unit that determines whether or not the number of people who eat and drink the meal information, and a plurality of the number of people who eat and drink the meal information include a food preference model corresponding to the individual identifier. And a meal information estimation unit that estimates the meal selected by the individual from the meal information and outputs the estimated meal information.

  According to the present invention, regarding the purchase history associated with one individual identifier, meal information included in the purchase history is consumed by the associated individual alone, or includes information on others. If other people's information is included, the information included in the purchase history is classified, and only the information consumed by the individual can be extracted and recorded.

  Furthermore, the purchase history classification system of the present invention, to the extent that each menu in the meal is eaten by the attribute information and past meal characteristics of the user who shared the meal, even when shared by multiple people There is an effect that it is possible to estimate, calculate a distribution ratio to each person, and automatically distribute meal information among users.

1 is a block diagram illustrating an example of a purchase history classification system according to a first embodiment of this invention. FIG. It is a figure which shows a 1st Example of this invention and shows an example of the user information which a user information management part manages. It is a figure which shows a 1st Example of this invention and shows an example of the meal information table which a meal information management part manages. It is a figure which shows a 1st Example of this invention and shows an example of the meal feature-value table which a meal feature-value management part manages. It is a flowchart which shows a 1st Example of this invention and shows an example of the process which applies and classifies a meal preference model to meal information. It is a flowchart which shows a 1st Example of this invention and shows an example of the process which shows the estimation result of meal information to a user. It is a flowchart which shows a 1st Example of this invention and shows an example of the process which produces | generates a food preference model. It is a screen image which shows the 1st Example of this invention and displays the estimation result of meal information. It is a block diagram which shows a 1st Example of this invention and shows an example of the functional element of a purchase history classification system. It is a block diagram which shows the 2nd Example of this invention and shows an example of the purchase history classification | category system. It is a figure which shows the 2nd Example of this invention and shows an example of the food preference degree table which a food preference management part manages. It is a figure which shows the 2nd Example of this invention and shows an example of the simultaneous eater setting information table which a simultaneous eater information management part manages. It is a figure which shows a 2nd Example of this invention and shows an example of the meal information table which a meal information management part manages. It is a flowchart which shows the 2nd Example of this invention and shows an example of the flow of a process which calculates food preference degree information. It is a flowchart which shows a 2nd Example of this invention and shows an example of the process which distributes meal information based on a meal preference degree. It is a screen image which shows the 2nd Example of this invention and displays the distribution result of meal information.

  Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

  Embodiment 1 of the present invention will be described in detail with reference to the drawings. In the first embodiment, information on meals taken by the user (or target person) is collected from a store or the like, and information on meals actually eaten by the user from the collected meal information is calculated from past history. A purchase history classification system that applies a model, extracts it, and records it in a database will be described.

  FIG. 1 is a block diagram showing an example of a purchase history classification system 100 according to the first embodiment of the present invention. The purchase history classification system 100 is composed mainly of a purchase history classification device 101 and a database 102.

  The purchase history classification system 100 is connected to a store information terminal 104 and a user terminal 110 via a network 103. The purchase history classification apparatus 101 includes an arithmetic device 108 that performs arithmetic processing, a memory 109 that stores programs and data, a storage medium 105 that stores programs and data, an input unit 106 that includes a keyboard, a mouse, a touch panel, and the like, It is a computer provided with the output part 107 containing a display.

  The storage medium 105 is a storage device that stores various programs, data, and the like that realize the purchase history classification system 100, and is, for example, a nonvolatile storage medium (magnetic disk drive, nonvolatile memory, or the like). The memory 109 is loaded with programs and data stored in the storage medium 105.

  The arithmetic device 108 is a device (processor) that executes a program loaded in the memory 109, and is a CPU, a GPU, or the like, for example. Processing and operations described below are executed by the arithmetic device 108.

  As a functional unit constituting the purchase history classification system 100, a meal information identifier (meal ID 301) is given when meal information is received from the store information terminal 104 via the network 103 and stored in the database 102. A meal information identifier setting unit 112 that performs the meal feature amount setting unit 111 that receives the setting of the meal feature amount of the meal provided by the store information terminal 104, and a meal feature that manages detailed feature amounts of the meal menu corresponding to the meal information An amount management unit 123; a utilization state setting unit 117 that sets a meal state such as the number of users; a meal trend analysis unit 113 that analyzes a meal trend from a past meal feature amount of the user; and a food preference based on the past meal trend A food preference model construction unit 114 for modeling the user, and a user who applies the food preference model to meal information including information for a plurality of people Meal information estimation unit 115 that estimates meal information actually eaten, a meal information update unit 119 that reconstructs meal information from the estimation result of the meal information estimation unit 115 and stores it in the database, and a user for the estimation result of the meal information A meal information correction input unit 116 that receives the correction request, and an accuracy setting unit 118 that calculates an accuracy 307 for the estimation result of the meal information.

  Each functional unit of the meal feature amount setting unit 111 to the meal information update unit 119 is loaded into the memory 109 as a program.

  The arithmetic device 108 operates as a functional unit that provides a predetermined function by processing according to a program of each functional unit. For example, the computing device 108 functions as the meal information identifier setting unit 112 by performing processing according to the meal information feature amount setting program. The same applies to other programs. Further, the arithmetic device 108 also operates as a function unit that provides each function of a plurality of processes executed by each program. A computer and a computer system are an apparatus and a system including these functional units.

  Information such as programs and tables for realizing each function of the purchase history classification apparatus 101 is a storage system, a nonvolatile semiconductor memory, a hard disk drive, a storage device such as an SSD (Solid State Drive), or an IC card, SD card, DVD Etc., and can be stored in a computer readable non-transitory data storage medium.

  The database 102 represents the feature of the meal information stored in the user information management unit 121 that manages attribute information for each user, the meal information management unit 122 that stores meal information for each user, and the meal information management unit 122. A meal feature amount management unit 123 that manages information and a food preference model management unit 124 that manages user-specific food preference models are included.

  The database 102 may be provided by a computer connected to the purchase history classification apparatus 101 via a network. Alternatively, the database 102 may be stored in a storage device connected to the purchase history classification device 101.

  The store information terminal 104 and the user terminal 110 are computers including an arithmetic device, a memory, an input device, and an output device (not shown). In addition, the store information terminal 104 includes a store meal feature amount management unit 1040 that manages a meal feature amount of meals provided in the store.

  FIG. 2 is a diagram illustrating an example of user information 200 managed by the user information management unit 121. The user information 200 includes a user ID 201 that identifies a user, a date of birth 202, a height 203, a weight 204, a gender 205, a job type 206, income information 207, and authentication ID information 208 in one record. Including. By using attributes such as gender 205 and job type 206 of this user information 1210, there is an effect that it becomes possible to categorize users (individuals) by group and build a food preference model.

  FIG. 3 is a diagram illustrating an example of the meal information table 300 managed by the meal information management unit 122. The meal information table 300 includes a user ID 201, a meal ID 301 that specifies meal information, a meal time 302 that indicates the generation time of the meal information, a use store 303 that specifies the store used, breakfast, lunch, dinner, and the like. A meal purpose 304 that indicates the time of the meal, a meal menu 305 that stores an identifier for specifying the menu of the meal that has been settled at the same time, a simultaneous eater information 306 that indicates the information of the person who has settled at the same time, The accuracy 307 representing the probability and the processed flag 308 are included in one record.

  The accuracy 307 is set to “1” when the meal information is accurate, and is set to “0” when the accuracy of the meal information is unknown. The initial value of the accuracy 307 is set to “0”.

  Further, the processed flag 308 is set to “True” when the record is processed in the estimation process described later, and is set to “False” when the record is not processed.

  As will be described later, the meal information table 300 has an effect that the user's actual meal information can be extracted using the meal preference model.

  FIG. 4 is a diagram illustrating an example of the meal feature amount table 400 managed by the meal feature amount management unit 123.

  The meal feature amount table 400 includes a provision store 401 that stores an identifier of a store that provides meals, a store business condition 402 that specifies conditions for providing a meal menu, a provision date 403, a menu ID 404 for a meal, and a menu name 405. , Price 406, nutritional value 407, ingredients 408, ingredients quantity 409, cooking method 410, cooking category 411, and menu category 412 are included in one record.

  In addition, about the foodstuff 408, the foodstuff quantity 409, and the cooking method 410, it can store about the information of several foodstuffs.

  By performing a predetermined process using the information in the meal feature table 400, there is an effect that the user's preference can be specified.

  Next, processing performed by the purchase history classification apparatus 101 will be described in detail using a flowchart.

  FIG. 5 is a flowchart showing a series of flows for classifying meal information. This process is executed each time the purchase history classification apparatus 101 receives new meal information.

  First, when a user settles a meal at a store, information on the meal is transmitted from the store information terminal 104 to the purchase history classification apparatus 101 (501). The purchase history classification apparatus 101 receives meal information, extracts user authentication ID information 208 included in the transmitted meal information, and refers to the user information 200 to identify the user ID 201.

  As the user authentication ID information 208, for example, an electronic money ID, a credit card, biometric authentication information, or other predetermined identification information can be used. The user authentication process is executed each time meal information is received.

  The user registers the authentication information in the purchase history classification apparatus 101 in advance, creates a user account, and acquires the user ID 201.

  The purchase history classification apparatus 101 assigns a meal information identifier (meal ID 301) to meal information for which the target user has been specified, and is stored in the meal information table 300 as the meal of the target user. At this time, the purchase history classification apparatus 101 sets initial values for the accuracy 307 and the processed flag of the meal information table 300. For example, the accuracy 307 is “0” and the processed flag is “false”. The meal ID 301 is given a unique value by the meal information identifier setting unit 112.

  Next, in step 502, the purchase history classification apparatus 101 refers to the processed flag 308 from the meal information table 300 of the meal information management unit 122, and acquires unprocessed (False) meal information. Note that the purchase history classification device 101 may specify the user ID 201 when extracting the data. This process is executed each time meal information is received.

  Next, in step 503, the purchase history classification apparatus 101 acquires a meal feature amount corresponding to the acquired meal information from the meal feature amount table 400 of the meal feature amount management unit 123, and calculates a meal feature amount for each meal cost menu. get. The purchase history classification apparatus 101 acquires a meal menu from the received meal information, and acquires a meal feature quantity from the meal feature quantity table 400 based on the menu ID 404 of the meal menu.

  The information on the meal feature table 400 is registered in advance in the meal feature table 400 of the purchase history classification apparatus 101 upon accepting registration from the store information terminal 104. The registration from the store information terminal 104 to the meal feature table 400 is, for example, input from the store information terminal 104 via the registration screen of the meal feature table 400 provided by the purchase history classification apparatus 101, and the store information. The terminal 104 may have a database (store meal feature amount management unit 1040) holding meal feature amounts, and may be transmitted from the database to the purchase history classification apparatus 101 and recorded in the meal feature amount management unit 123. Note that the meal feature value setting unit 111 executes reception and registration of the meal feature value.

  Next, the purchase history classification apparatus 101 determines in step 504 whether the target meal information includes not only the target user alone but also information eaten by a plurality of persons. This processing is performed by the usage status setting unit 117 shown in FIG.

  The purchase history classification apparatus 101 refers to the simultaneous eater information 306 of the meal information to be processed, and acquires the value of the simultaneous eater information 306 transmitted by the store information terminal 104. The purchase history classification apparatus 101 determines that there is a simultaneous eater when the value of the simultaneous eater information 306 is 2 or more, and proceeds to the processing of step 505. On the other hand, if there are no simultaneous eaters, the process proceeds to step 507. This process determines whether or not a plurality of eaters are included in a fare paid by one user.

  Next, in step 505, the purchase history classification apparatus 101 acquires the target user's food preference model from the food preference model management unit 124 when there is simultaneous eater information. The purchase history classification apparatus 101 inputs the meal feature amount of each meal menu to the parameters of the acquired meal preference model, calculates the selection probability of each meal menu 305 of the meal information table 300, and the selection probability is equal to or greater than a separately determined threshold value. Extract the meal menu. The purchase history classification apparatus 101 generates an estimated meal menu (estimated meal information) after eliminating patterns that are difficult to combine in the past meal history with respect to the extracted menu. This process is performed by the meal information estimation unit 115 shown in FIG.

  Next, in step 506, the purchase history classification apparatus 101 classifies the relationship between the plurality of menus included in the meal information table 300 and the eater from the information of the estimated eating menu generated in step 505. In other words, the purchase history classification device 101 classifies a menu that is actually eaten or eaten as a eater based on the user's estimated eating menu. This process may be performed only for the user currently focused on.

  Next, purchase history classification apparatus 101 sets a value of accuracy 307 in meal information table 300 in step 307. For example, when the user has not confirmed the estimation information, the purchase history classification apparatus 101 sets the accuracy 307 as “0”. If the user has confirmed the estimated information, the accuracy 307 is set to “1”. This process is performed by the accuracy setting unit 118 shown in FIG.

  After the setting of the accuracy 307 is completed, the purchase history classification apparatus 101 sets the processed flag 308 to “True” and stores the meal information in the meal information management unit 122. This process is performed by the meal information update unit 119 shown in FIG.

  By the above processing, the purchase history classification apparatus 101 estimates the menu of the eater from the meal preference model when the eater information is received from the store information terminal 104 and includes a plurality of eaters. In other words, the purchase history classification apparatus 101 estimates the menu of the eater from the food preference model when a plurality of eaters are included in the settlement of one user. Then, the purchase history classification apparatus 101 classifies the menu as a eater based on the estimation result.

  FIG. 6 is a flowchart illustrating an example of a probability setting process in which the purchase history classification apparatus 101 presents meal estimation information to the user terminal 110 and receives confirmation or correction of the content of the estimation information from the user terminal 110. Executing this process when a request from the user terminal 110 or the like is received corresponds to the accuracy setting unit 118 in FIG. Further, this process may be executed in step 507 of FIG.

  First, in step 602, the purchase history classification apparatus 101 transmits the eating menu estimation result to the user terminal 110 that has received the request.

  The user can check the estimation result of the eating menu with an output device (not shown) of the user terminal 110 connected via the network 103.

  It is assumed that the user operating the user terminal 110 confirms the estimation result without correction, the case where the user confirms after the correction, and the case where the estimation result is neither corrected nor approved. First, the purchase history classification apparatus 101 determines whether a request for finalizing the estimation result has been received from the user terminal 110 (603). If a request to confirm is accepted, the process proceeds to step 606; otherwise, the process proceeds to step 604.

  Next, purchase history classification apparatus 101 determines in step 604 whether or not a correction request for an estimation result has been received from user terminal 110. If the purchase history classification apparatus 101 accepts a correction request, the process proceeds to step 605; otherwise, the process proceeds to step 606.

  Next, purchase history classification apparatus 101 accepts correction of the estimation result from user terminal 110 at step 605.

  Next, the purchase history classification device 101 sets the accuracy 307 of the meal information table 300 high when the confirmation or correction by the user is received from the user terminal 110 in step 606. The purchase history classification apparatus 101 increases the value of the accuracy 307 and updates the value to “1” when the initial value is “0”, for example.

  Next, the purchase history classifying apparatus 101 updates the contents of the corresponding meal information in the meal information table 300 stored in the meal information management unit 122 in the meal information update step 607 or the correction contents received from the user terminal 110 or The meal information and the accuracy information 307 are updated by reflecting the confirmed contents.

  Through the above processing, the eating menu estimated by the purchase history classification apparatus 101 is confirmed or corrected by the user terminal 110, and the meal information table 300 is updated.

  FIG. 7 is a flowchart illustrating an example of processing in which the purchase history classification apparatus 101 constructs a food preference model. This process is executed in response to a command from an administrator of the purchase history classification apparatus 101 or a predetermined cycle (for example, every week) (701). The following process shows an example of generating a food preference model for one user. Further, this process corresponds to the food preference model construction unit 114 shown in FIG.

  First, in step 702, the purchase history classification apparatus 101 selects meal information to be used for generating a food preference model from past meal information accumulated in the meal information table 300. This selection is performed using the user ID 201, the meal purpose 304, the accuracy 307, and the meal time 302 of the meal information table 300.

  For example, for the purpose of building a model relating to lunch, the purchase history classification apparatus 101 designates the user ID 201, the meal purpose 304 is “lunch”, and the accuracy 307 is equal to or greater than a certain value (for example, “1”). Select only meal information. Note that a food preference model with high accuracy can be generated by selecting meal information having a large value of the accuracy 307.

  Next, in step 703, the purchase history classification apparatus 101 selects the meal feature amount of the meal information selected in step 702 from the meal feature amount table 400 of the meal feature amount management unit 123, and is necessary for generating a food preference model. To extract various meal features.

  For example, the purchase history classification apparatus 101 has a menu name 405, a price 406, a menu category 412 corresponding to the eating menu 305, a menu number, and a nutrition intake (nutrient value 407) for each meal time 302 and meal purpose 304 of the meal information. To get. Note that not only a single menu but also a pattern for selecting a plurality of menus can be acquired as the feature amount.

  Then, the purchase history classification apparatus 101 calculates statistical values such as variance and deviation with respect to the acquired value (feature amount), and analyzes the selection tendency of the meal (menu) of the user ID 201. This process is performed by the meal tendency analysis unit 113 shown in FIG. Note that calculation of statistical values such as variance and deviation may be performed using a known or well-known method, and thus will not be described in detail here.

  In addition, the statistical values calculated by the purchase history classification apparatus 101 include individual frequency information such as the selection frequency of the meal menu (305, 404, 405) selected by the target user ID 201 in the past, and the food 408 included in each meal menu. The index representing the preference can be a meal selection tendency.

  Next, in step 704, the purchase history classification apparatus 101 creates a model for calculating the selection probability of the meal menu (404, 405) included in the meal feature amount table 400 based on the selection tendency of the meal menu. . This process is performed by the food preference model construction unit 114 shown in FIG. 1, and the food preference model of the user ID 201 that is the current target is calculated.

  The model for calculating the menu selection probability can generate a prediction (meal preference) model for predicting the selection probability of the meal menu using the feature amount extracted in step 703 as an explanatory variable. Then, the purchase history classification apparatus 101 stores the generated meal menu selection probability in the food preference model management unit 124 as a food preference model.

  Since the user's food preference changes depending on the year, environment, or season, it is necessary to update the food preference model as appropriate, but if the above process is repeated when new meal information is accumulated, A highly accurate food preference model can be constructed.

  The meal information used to generate the meal preference model may be set to a period such as one year or several months.

  Moreover, although the example which produces | generates the meal preference model regarding one meal objective 304 was shown about one user in the said process, you may produce | generate a meal preference model for every several meal objective 304 about one user. . For example, for one user, the meal purpose 304 generates three meal preference models of “breakfast”, “lunch”, and “dinner”, thereby obtaining a meal menu selection probability corresponding to the meal purpose 304. be able to.

  By performing the above process for all users, a food preference model for each user can be generated and held in the food preference model management unit 124.

  FIG. 8 is a screen image of a meal information estimation result that the purchase history classification apparatus 101 transmits to the user terminal 110. A screen 801 for displaying the estimation result of the meal information is transmitted to the user terminal 110 in step 602 of FIG. 6, and the purchase history classification apparatus 101 receives user confirmation and correction from the user terminal 110.

  The screen 801 for displaying the estimation result of the meal information includes a meal condition display unit 802, an estimated breakdown display unit 803, an estimation result display unit 804, a confirmation button 805, and a correction button 806.

  The meal condition display unit 802 displays the store name, date and time, and purpose. These pieces of information include a meal time 302, a use store 303, and a meal purpose 304 in the meal information table 300. The meal condition display unit 802 indicates the date and place of eating the displayed menu.

  The estimated breakdown display unit 803 indicates the selection probability of each menu calculated by the food preference model. The estimation result display unit 804 shows the result of the estimated meal information (menu).

  The user operates the input device of the user terminal 110 and clicks the confirmation button 805 to confirm the estimation result. When the correction button 806 is clicked, the user can call a function for correcting the estimation result.

  FIG. 9 is a block diagram illustrating an example of functional elements of the purchase history classification system 100. FIG. 9 shows an outline of processing performed in the purchase history classification system 100.

  The purchase history classification apparatus 101 stores the meal information history adjusted by each user in the meal information table 300. The purchase history classification apparatus 101 executes the food preference model construction unit 114 shown in FIG. 7 at a predetermined cycle or the like to generate a food preference model 1240 of the target user.

  In the generation of the meal preference model 1240, the purchase history classification apparatus 101 acquires meal information within a predetermined period from the meal information table 300 for each user ID 201 and meal purpose 304 (702).

  The purchase history classification apparatus 101 acquires a meal feature amount from the meal feature amount table 400 for the selected meal information, and analyzes the selection tendency of the meal menu (703). The purchase history classification apparatus 101 calculates the probability that a pre-registered meal menu is selected based on the selection tendency of the meal menu, and generates a meal preference model 1240 (704).

  Upon receiving new meal information (502), the purchase history classification apparatus 101 reads the user's meal preference model 1240 and calculates the meal menu selection probability when the meal information includes a plurality of eaters. The meal menu selected by the user is estimated and output to the user terminal 110 as an estimation result (602).

  In the user terminal 110, when the user confirms or corrects the estimation result of the meal menu, the meal information accuracy 307 is set to “1”, otherwise, “0” is set and the meal information is updated. To do.

  As described above, in the purchase history classification system 100, it is possible to estimate a meal menu selected by a corresponding user from meal information including a plurality of eaters using a meal preference model 1240 generated in advance.

  Moreover, the estimation accuracy of a meal menu can be improved by using meal information with high accuracy 307 (above a threshold) confirmed or corrected by the user in the history of meal information.

  In the first embodiment, a probability model having the meal selection probability as the explanatory variable is used as the meal selection probability. However, as a method for calculating the selection probability, any model that can calculate the meal selection probability may be used. For example, a logistic regression model or a Bayesian network may be used.

  Moreover, in the meal information selection process of step 702 shown in FIG. 7, selection according to the target user and the meal purpose 304 is performed. However, the selection is not limited to this. A food preference model can be created. In addition, if an arbitrary category such as age, gender 205, job type 206, income information 207, etc. is selected as a condition, there is an effect that it is possible to create a food preference model for each category instead of for each individual.

  Moreover, in the said Example 1, although the value of the simultaneous eater information 306 is acquired from the meal information table 300 of the meal information management part 122, the case where the value of the simultaneous eater information 306 is missing is also considered. However, even if the simultaneous eater information is not recorded, the simultaneous eater based on the past information recorded in the meal information management unit 122 and the meal feature amount management unit 123 in step 504 of FIG. The presence or absence of can be determined. For example, determination standard information such as an average price and the number of menus per target user is calculated, and based on this, it is determined whether or not the target meal information includes information for one person or more. Moreover, using the information of the nutritional value 407 as a determination criterion, a threshold value per person may be provided for the nutrition intake taken by the target user, and the presence or absence of a simultaneous eater may be determined.

  Embodiment 2 of the present invention will be described in detail with reference to the drawings. In the second embodiment, a purchase history classification system that divides meal information of a corresponding menu and distributes it to each user when a single menu (for example, a serving menu etc.) is shared (or distributed) by a plurality of eaters. explain.

  FIG. 10 is a block diagram illustrating an example of a purchase history classification system 100 that is Embodiment 2 of the present invention. The purchase history classification system 100 is composed mainly of a purchase history classification apparatus 101 and a database 102. The purchase history classification apparatus 101 is connected to the store information terminal 104 and the user terminal 110 via the network 103. The purchase history classification device 101 is a computer device, and includes an input unit 106 that receives input of various types of information, an output unit 107 including a display, an arithmetic unit 108, a memory 109, and a storage medium 105.

  Further, in addition to the configuration of the first embodiment, the purchase history classification apparatus 101 extracts meal feature quantities from meal information histories, and the ingredients 408 and cooking methods 410 that are closely related to meal preferences from the meal feature quantities. A meal preference level calculation unit 901 that calculates a user's preference level using feature quantities such as the same, and a meal that is shared and ordered by a plurality of people is allocated to each user according to a meal preference level A distribution coefficient calculation unit 902 that calculates a distribution coefficient for the user, and a simultaneous eater setting unit 903 that sets information of a user who shared a meal using a template or the like.

  Further, in addition to the configuration of the first embodiment, the database 102 manages the preference information management unit 904 that manages the preference information for each user and the setting information used when setting the information of the simultaneous eaters. And a simultaneous eater information management unit 905.

  In the second embodiment, by using the preference information and the setting information of the simultaneous eater information, there is an excellent effect that the meal information can be distributed even when the meal menu is shared by a plurality of people. Is.

  FIG. 11 is a diagram illustrating an example of a food preference table 1000 managed by the food preference information management unit 904.

  The meal preference level table 1000 includes a user ID 201 for identifying a user, a meal purpose 1001 for managing preferences for each meal purpose 304, simultaneous eater information 1002 for managing preferences for each simultaneous eater, and a meal menu. One record includes an ingredient preference degree 1003 for storing the preference degree for the contained ingredient (material) for each ingredient, and a cooking method preference degree 1004 for storing the preference degree for the cooking method of the menu for each cooking method.

  With this information, there is an effect that the user's food preference can be finely managed for each user's meal purpose 304.

  FIG. 12 is a diagram illustrating an example of the simultaneous eater setting information table 1100 managed by the simultaneous eater information management unit 905.

  The simultaneous eater setting information table 1100 stores a user ID 201 for identifying a user and a user ID belonging to a workplace colleague (setting 1), family (setting 2), friend (setting 3), etc. as a unit of a group to eat at the same time. The simultaneous eater setting information 1101 to be included is included in one record.

  In the example shown in the drawing, an example is shown in which the simultaneous eater setting information 1101 is divided into three groups and set to setting 1 to setting 3. However, the present invention is not limited to this.

  The simultaneous eater setting information table 1100 has an effect that a person who eats with the corresponding user can be easily set.

  FIG. 13 is a diagram illustrating an example of a meal information table 300 managed by the meal information management unit 122.

  In addition to the configuration of the first embodiment, the meal information table 300 according to the second embodiment includes shared menu information 311 that is menu information that the user has shared and ingested with others, and sharer information 312 that indicates a shared partner. including.

  The shared menu information 311 and the sharer information 312 are set when being transmitted from the store information terminal 104 to the purchase history classification apparatus 101 system or from the user terminal 110 via an input screen (not shown). By adding the share menu information 311 and the sharer information 312, there is an effect that it becomes possible to specify the shared meal information and the sharer.

  Next, processing performed by the purchase history classification apparatus 101 will be described in detail using a flowchart.

  FIG. 14 is a flowchart illustrating an example of a process in which the purchase history classification apparatus 101 creates meal preference information. This process is executed in accordance with a command from the administrator of the purchase history classification apparatus 101 or a predetermined cycle (for example, every week) (1301). The following processing shows an example of generating the food preference table 1000 for one user. Further, this processing corresponds to the meal preference level calculation unit 901 illustrated in FIG.

  In step 1302, the purchase history classification apparatus 101 filters the meal information used to create the food preference information from the past meal information accumulated in the meal information table 300 with the accuracy 307, for example, the accuracy 307 Only meal information whose value is greater than or equal to a predetermined value is selected.

  Next, the purchase history classification apparatus 101 acquires the meal feature amount of the selected meal information from the meal feature amount table 400 of the meal feature amount management unit 123 in step 1303.

  Next, in step 1304, the purchase history classification apparatus 101 calculates the preference level of the target menu based on the meal feature amount. By quantifying the degree of preference as a common index for each user, it is possible to distribute meal information according to the degree of preference among users who share a menu.

  As an example of the preference degree calculation method, the appearance frequency of a certain meal feature amount with respect to the number of meals can be used. For example, when meat is taken in all of the meals for one month, the “meat” column is set in the food preference 1003 and the preference value is set to 1. In addition, a “baked” column is set as the cooking method preference degree 1004, and a preference value is set. The preference degree calculation process is updated, for example, every week in order to reflect a change in the user's preference. However, the update frequency can be arbitrarily changed.

  Next, the purchase history classification apparatus 101 stores the information (1003, 1004) of the target user's food preference level in the food preference level information management unit 904 in the preference level information storage step 1305.

  In the above-described processing, an example in which the food preference level is calculated for one user with respect to one meal purpose 1001 is shown, but the food preference level may be calculated for each of a plurality of meal purposes 1001 for one user. For example, by calculating the three food preference levels of “breakfast”, “lunch”, and “dinner” for one user, the distribution coefficient of the meal menu corresponding to the meal purpose 1001 can be accurately calculated. Can get to.

  By performing the above process for all users, the food preference level for each user can be calculated and held in the food preference level information management unit 904.

  FIG. 15 is a flowchart illustrating an example of a process for allocating a menu of meal information based on the meal preference table 1000. Steps 501 to 508 in FIG. 15 are the same as those in the first embodiment, and the purchase history classification apparatus 101 acquires unprocessed data whose processed flag 308 is “False” from the meal information table 300. A meal feature amount corresponding to the data is acquired. And purchase history classification device 101 presumes the user's meal menu using a meal preference model, when there is a person who eats at the same time.

  Next, purchase history classification apparatus 101 determines in step 1401 whether or not a certain meal menu is shared with other eaters. Whether or not the meal menu is shared can be determined based on the contents of the simultaneous eater setting information 1101 recorded in the sharer information 312 shown in FIG.

  Next, in step 1402, the purchase history classification apparatus 101 acquires the food preference level table 1000 from the food preference level information management unit 904 based on the sharer information 312. Calculate the distribution ratio.

  The purchase history classification apparatus 101 acquires the food preference level related to the meal feature amount such as the ingredients 408 and the cooking method 410 for the meal menu to be processed for the user set in the sharer information 312, and the food preference The distribution coefficient is calculated by comparing the degree values.

  For calculating the distribution coefficient, a predetermined meal feature amount serving as a reference can be selected. For example, if the feature amount relating to the ingredients and the feature amount relating to the cooking method are used, and the ingredient 408 of the menu to be processed is the ingredient A and the cooking method 410 is the recipe A, the ingredient preference level 1003 indicates the ingredient A. The preference level is acquired, and the preference level of the cooking method A is acquired from the cooking method preference level 1004.

  The purchase history classification apparatus 101 maps a vector of preference levels created based on the preference value to a vector space with the preference level of the selected meal feature quantity as an axis, and calculates a vector norm. The distribution coefficient is obtained by calculating the norm of a user's preference degree vector with respect to the total vector norm of each user. The distribution coefficient is calculated as a value between 0 and 1. If the past meal information of the simultaneous eater is not recorded, the input of the profile information of the simultaneous eater is accepted, and for example, it is calculated by substituting the preference level of the existing user with the same age, gender 205, etc. Good.

  In step 1403, the purchase history classification apparatus 101 calculates a distribution ratio for each user based on the product of the calculated distribution coefficient and the menu nutrition value 407, and distributes the obtained value. Next, in step 507, the purchase history classification apparatus 101 sets the accuracy 307 as in the first embodiment, and stores the meal information in the meal information table 300 of the meal information management unit 122.

  FIG. 16 is a screen image of a meal information distribution result that the purchase history classification apparatus 101 transmits to the user terminal 110.

  The purchase log | history classification apparatus 101 shows an example of the screen image which presents the distribution result of the meal information shared between users to the user terminal 110, and receives a user's confirmation and correction.

  A screen 1501 for displaying the distribution result of meal information includes a meal condition display unit 1502, a distribution breakdown display unit 1503, a distribution result display unit 1504, a confirmation button 1505, a correction button 1506, and a simultaneous eater setting unit 1507. , A detailed setting button 1508 and a distribution recalculation button 1509.

  The meal condition display section 1502 shows the place where the meal was taken, the meal time, and the purpose of the meal, and the distribution breakdown display section 1503 shows the distribution coefficient of each meal menu calculated by the food preference table 1000. The distribution result display unit 1504 indicates the result of the distributed meal information.

  In the user terminal 110, a distribution result can be confirmed by clicking a confirmation button 1505, and a function for correcting the distribution result can be called by pressing a correction button 1506.

  The simultaneous eater setting unit 1507 can set or change a template of a preset group of simultaneous drinkers. In the case of performing detailed editing of a simultaneous eater, a setting function can be called by clicking a detailed setting button 1508. When the distribution recalculation button 1509 is pressed, the distribution can be recalculated using the simultaneous eater information whose setting has been changed.

  Note that the screen 1501 may be transmitted to the user terminal 110 in step 1403 of FIG.

  In the second embodiment, the example of using the frequency for each meal feature amount is shown as the food preference level. However, as long as information indicating the food preference level can be obtained, it may be calculated using another statistical method. .

  As described above, in the second embodiment, the degree of food preference for each user is calculated for each ingredient and cooking method from the accumulated meal information table 300. When a single meal menu is shared by a plurality of eaters, the meal menu consumed by each of the plurality of users is calculated as the distribution result by calculating the distribution coefficient from the food preference of the ingredients and the cooking method. be able to. Thereby, even when one dish is divided by a plurality of eaters, it is possible to grasp the amount of the meal menu taken by each user.

  The configuration of the computer, the processing unit, the processing unit, and the like described in the present invention may be partially or entirely realized by dedicated hardware.

  In addition, the various software exemplified in the present embodiment can be stored in various recording media (for example, non-transitory storage media) such as electromagnetic, electronic, and optical, and through a communication network such as the Internet. It can be downloaded to a computer.

  The present invention is not limited to the above-described embodiments, and includes various modifications. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described.

Claims (12)

A purchase history classification system comprising: a computer including a processor and a memory; and a terminal connected to the computer via a network,
The calculator is
A meal information storage unit for receiving and accumulating meal information including an identifier for identifying an individual and an identifier of a meal eaten or consumed by the individual;
A meal feature amount storage unit for storing the feature amount of the meal received from the terminal;
For each individual identifier, the meal information is selected from the meal information storage unit, the feature amount corresponding to the meal information is acquired from the meal feature amount storage unit, and the tendency of the individual meal selection is analyzed. A food trend analysis unit
A food preference model construction unit that calculates the meal selection probability as a food preference model based on the analysis result of the tendency of the meal selection for each individual;
An input unit for receiving input of the meal information;
A determination unit for determining whether or not the number of people eating and drinking in the meal information is plural;
When the number of people who eat and drink in the meal information is plural, the meal selected by the individual is estimated from the meal information based on the meal preference model corresponding to the identifier of the individual, and is output as estimated meal information An information estimation unit;
A purchase history classification system characterized by comprising: The purchase history classification system according to claim 1,
The meal information is
Including the number of people eating and drinking at the same time,
The determination unit
When there is a person who eats and drinks at the same time, it is determined that the number of people who eat and drink is plural based on the meal information. The purchase history classification system according to claim 1,
The determination unit
A purchase history classification system for determining whether or not there are a plurality of people who eat and drink based on a meal feature amount corresponding to the meal information and a preset criterion. The purchase history classification system according to claim 1,
A meal information correction input unit that receives corrections for the estimated meal information estimated by the meal information estimation unit;
An accuracy setting unit that accepts confirmation of the estimated meal information and sets the accuracy of the meal information;
A purchase history classification system characterized by further comprising: The purchase history classification system according to claim 1,
The meal information is
Including shared menu information for identifying meals shared by the plurality of people, and an identifier for identifying individuals who have shared the shared menu information,
A food preference information storage unit that stores the preference calculated for each individual based on the meal information;
A distribution coefficient calculation unit that calculates a distribution coefficient for distributing the meal information of the shared menu information to a plurality based on the preference degree;
A purchase history classification system characterized by further comprising: The purchase history classification system according to claim 5,
The individual meal information stored in the meal information storage unit and the meal feature amount corresponding to the meal information are acquired, the preference degree is calculated in advance based on the appearance frequency of the meal feature amount, and the meal is stored. A purchase history classification system, further comprising a food preference level calculation unit stored in the preference level information storage unit. A purchase history classification method for classifying a purchase history comprising a computer having a processor and a memory, and a terminal connected to the computer via a network,
A first step in which the calculator receives meal information including an identifier for identifying an individual and an identifier of a meal eaten or consumed by the individual from the terminal and accumulates the meal information in a meal information storage unit;
A second step in which the calculator stores the feature quantity of the meal received from the terminal in a meal feature quantity storage unit;
The computer selects the meal information from the meal information storage unit for each individual identifier, acquires the feature amount corresponding to the meal information from the meal feature amount storage unit, and selects the individual meal A third step of analyzing the trend of
A fourth step in which the calculator calculates the meal selection probability as a meal preference model based on the analysis result of the tendency of the meal selection for each individual;
A fifth step in which the calculator receives an input of the meal information and determines whether or not the number of people who eat and drink the meal information is plural;
When the computer has a plurality of people who eat and drink the meal information, the meal selected by the individual is estimated from the meal information based on the meal preference model corresponding to the individual identifier, and the estimated meal information As a sixth step,
A purchase history classification method comprising: The purchase history classification method according to claim 7,
The meal information is
Including the number of people eating and drinking at the same time,
The fifth step includes
When there is a person who eats and drinks at the same time, it is determined that the number of people who eat and drink is plural based on the meal information. The purchase history classification method according to claim 7,
The fifth step includes
A purchase history classification method, comprising: determining whether or not there are a plurality of people who eat and drink based on a meal feature amount corresponding to the meal information and a preset determination criterion. The purchase history classification method according to claim 7,
A seventh step in which the calculator receives a correction for the estimated estimated meal information;
An eighth step in which the calculator accepts confirmation for the estimated meal information and sets the accuracy of the meal information;
The purchase history classification method characterized by including further. The purchase history classification method according to claim 7,
The meal information is
Including shared menu information for identifying meals shared by the plurality of people, and an identifier for identifying individuals who have shared the shared menu information,
A ninth step in which the calculator stores a preference degree calculated for each individual based on the meal information in a meal preference degree information storage unit;
A tenth step in which the calculator calculates a distribution coefficient for distributing the meal information of the shared menu information to a plurality based on the degree of preference;
The purchase history classification method characterized by including further. The purchase history classification method according to claim 11,
The calculator acquires meal information for each individual stored in the meal information storage unit and the meal feature amount corresponding to the meal information, and calculates the preference degree in advance based on the appearance frequency of the meal feature amount. The purchase history classification method further comprising an eleventh step of storing in the food preference information storage unit.

Images