JP2017211852A - Advice support system and advice support method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable advices to be made so as to allow many users to maintain their motivation.SOLUTION: A server 10 comprises: a user information input unit 12 that inputs user information on a user itself from a user; an advice candidate creation unit 13 that creates candidates of advices with respect to the user on the basis of the input user information; an advice candidate output unit 14 that outputs the created candidates of the advices with respect to supporters; and advice input unit 15 that inputs the advice in accordance with the output advice from the supporter; and an advice output unit 16 that outputs the input advice to the user.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an advice support system and an advice support method for supporting advice on health and the like for a user.

  As shown in Patent Document 1, a system has been proposed in which a user and a registered dietitian communicate with each other through sharing meal information and exchanging comments. Through this system, the registered dietitian provides dietary guidance to the user.

Japanese Patent Laying-Open No. 2015-191470

  As described above, a mechanism for remote nutrition guidance or health consultation has been conventionally provided. However, the number of medical staff including doctors and registered dietitians who can engage in nutritional guidance or health consultation is limited. Therefore, it is difficult for those medical staff to give advice to the majority of users who need advice.

  In order to give advice to the majority of users, it is conceivable that the system automatically gives advice. However, the automatic advice provided by the system lacks the elements of communication, which are essential elements such as nutritional guidance, and has little effect on maintaining motivation when compared to people. Guidance techniques such as nutritional guidance are an important factor for improving lifestyle habits, but for users aiming for a healthy life at the same time, maintaining motivation to achieve a healthy life is the biggest challenge. It is. No efforts have been made so far, such as the above-mentioned conventional techniques, to get plenty of encouragement from the surroundings and advice on health.

  The present invention has been made in view of the above, and an object of the present invention is to provide a management system and a management method that enable many users to give advice so that motivation can be maintained.

  In order to achieve the above object, the advice support system according to the present invention is input by user information input means for inputting user information about the user from the advised user who is the user who receives advice, and user information input means. Advice candidate generation means for generating advice candidates for the advised user based on user information, and advice candidates for outputting advice candidates generated by the advice candidate generation means to advice users who are advice users The advice input means for inputting advice according to the advice candidate output by the output means and the advice candidate output means from the advice user, and the advice input by the advice input means Comprising an advice output means outputs to Za, a.

  The advice support system according to the present invention can present advice candidates to the advice user. Thereby, a general user who is not an expert such as a registered dietitian can be an advice user. Thereby, advice between users can be activated. That is, according to the advice support system according to the present invention, it is possible to provide advice so that many users can maintain motivation.

  The user information input unit inputs text as user information, and the advice candidate generation unit extracts information indicating food from the text input by the user information input unit, and selects an advice candidate based on the extracted information. It may be generated. According to this structure, suitable advice based on a user's meal is performed between users, for example.

  The advice candidate generation means may calculate an amount for each nutritional component based on the extracted information, and generate an advice candidate based on the calculated amount. According to this configuration, appropriate advice based on the nutritional components of the user's meal is provided between the users.

  The user information input means may input information indicating user characteristics as user information. According to this configuration, for example, appropriate advice based on the personality of the user is performed between the users.

  The advice input means inputs advice that has been corrected for the advice candidate, and the advice candidate generation means is also based on the advice that has been input by the advice input means before generation of the advice candidate. The advice candidates may be generated. According to this configuration, the advice candidate can be corrected to the advice user, and an appropriate advice candidate can be presented according to the corrected advice.

  The advice support system further includes response input means for inputting a response from the advised user according to the advice output by the advice output means, and the advice candidate generation means is input by the response input means before generation of the advice candidate. The advice candidate may be generated on the basis of the received response. According to this configuration, an appropriate advice candidate can be presented in accordance with a response from the advised user.

  By the way, the present invention can be described as an invention of an advice support method as described below, as well as an invention of an advice support system as described above. This is substantially the same invention only in different categories, and has the same operations and effects.

  That is, the advice support method according to the present invention is an advice support method that is an operation method of an advice support system, and a user information input step for inputting user information about itself from a user to be advised who is a user who receives advice; Based on the user information input in the user information input step, an advice candidate generation step for generating advice candidates for the advised user, and an advice user who provides advice on the advice candidates generated in the advice candidate generation step Advice candidate output step to be output to the user, and advice input step to input advice according to the advice candidate output in the advice candidate output step from the advice user When, including advice input in advice input step, and advice output step of outputting against the advice the user, the.

  According to the present invention, a general user who is not an expert such as a registered dietitian can be an advice user. Thereby, advice between users can be activated. That is, according to the present invention, it is possible to provide advice so that many users can maintain motivation.

It is a figure which shows the structure of the server which is the advice assistance system which concerns on embodiment of this invention. It is a figure which shows the information memorize | stored in a memory | storage part. It is a figure which shows the example of the screen displayed in a client terminal. It is a figure which shows the example of the screen displayed in a client terminal. It is a figure which shows the example of the screen displayed in a client terminal. It is a figure which shows the specific example of learning data and transmission advice candidate data. It is a flowchart which shows the process performed when a comment is input from a user with the server which is the advice assistance system which concerns on embodiment of this invention. It is a flowchart which shows the process performed when the candidate which is an advice is output to a supporter at the server which is the advice assistance system which concerns on embodiment of this invention, and an advice is input from a supporter. It is a flowchart which shows the process performed when the server which is the advice assistance system which concerns on embodiment of this invention outputs advice to a user. It is a flowchart which shows the process performed when the candidate of an advice is updated with the server which is the advice assistance system which concerns on embodiment of this invention.

  Hereinafter, embodiments of an advice support system and an advice support method according to the present invention will be described in detail with reference to the drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 shows a server 10 that is an advice support system according to the present embodiment. The server 10 is a device that manages information related to user health. In particular, the server 10 manages information about the user's meal as information about the user's health. In addition, the server 10 may manage information related to exercise and information measured by the user such as weight. The user accesses the server 10 using the client terminal 20 and registers information related to health in the server 10. The user can use the function of the server 10 for own health management.

  Through the server 10, the user can give advice on health to another user by referring to the information of the other user. Specifically, the user registers user information about himself / herself to be disclosed to another user in the server 10. For example, as user information, the user registers a comment, which is a text (sentence) indicating his / her meal content, in the server 10. The user refers to a comment registered by another user and gives advice according to the comment. For example, when a user registers a comment saying, “I ate ramen and gyoza today. Are you eating too much?”, Another user responded, “You should eat vegetables. "Be careful not to eat too much carbs. Let's do our best tomorrow!"

  In the server 10 according to the present embodiment, advice candidates are presented when advice is given to another user. Therefore, when giving advice to another user, specialized knowledge about health is not necessarily required. Thereby, the advice between users is performed actively. This is a user's motivation, and health management using the function of the server 10 is continuously performed.

  In the server 10, all users registered in the server 10 can be set as advised users who receive advice. Alternatively, only some preset users may be set as advised users. In the server 10, all users registered in the server 10 can be advice users who are advice users. Alternatively, only some preset users may be advised users. In this embodiment, the user to be advised is simply called a user, and the advice user is called a supporter (supports another user by giving advice).

  The server 10 is realized by a computer such as a PC (Personal Computer) or a server device including hardware such as a CPU (Central Processing Unit), a memory, and a communication module. The functions of the server 10 to be described later are exhibited by operating these components by a program or the like. The server 10 may be realized by a computer system including a plurality of devices.

  The client terminal 20 is a device for using the function of the server 10 and is used by a user. The client terminal 20 has a function of transmitting / receiving information to / from the server 10 and a function of inputting / outputting information. As the client terminal 20, a conventional mobile terminal such as a smartphone or a PC is used. The server 10 and the client terminal 20 can exchange information with each other via a communication network such as the Internet.

  Subsequently, functions of the server 10 according to the present embodiment will be described. As shown in FIG. 1, the server 10 includes a storage unit 11, a user information input unit 12, an advice candidate generation unit 13, an advice candidate output unit 14, an advice input unit 15, an advice output unit 16, and a response input. And a unit 17.

  The storage unit 11 is a storage unit that stores information necessary for processing in the server 10. Information stored by the storage unit 11 is shown in FIG. What kind of information is used and how it is used will be described later.

  The user information input unit 12 is user information input means for inputting user information about itself from the user. The user information input unit 12 inputs text as user information.

  The text is, for example, a comment that is a text (sentence) indicating the content of the meal described above. The user information input unit 12 receives and inputs a comment transmitted from the user's client terminal 20. A comment is transmitted from the client terminal 20 in response to a user operation on the client terminal 20.

  For example, when the user's client terminal 20 accesses the server 10, the client terminal 20 displays the user top screen shown in FIG. When the button B1 displayed as “input” is pressed by the user's operation on the user top screen, a screen for inputting a comment shown in FIG. 3B is displayed. This screen is provided with a form F1 for inputting a comment. When a comment is input to the form F1 by the user's operation and the button B2 displayed as “Save” is pressed, the comment is transmitted from the client terminal 20 to the server 10. At this time, information other than the comment, for example, an image of eaten food may be transmitted together.

  Each user is given a user ID, which is information for identifying the user, in advance. The user ID is also information for specifying a supporter. When a comment is transmitted from the client terminal 20 to the server 10, it is transmitted together with the user ID. The user information input unit 12 stores the received comment in the storage unit 11 as advice waiting data shown in FIG. As shown in FIG. 2A, the advice waiting data is information in which a comment ID, a user ID, a posted comment, and the like are associated with each other. The comment ID is information for specifying a comment input by the user, and is given, for example, when the user information input unit 12 stores advice waiting data in the storage unit 11. The user ID and the posted comment are the user ID and the comment received from the client terminal 20.

  The user information input unit 12 inputs information indicating user characteristics as user information. The information indicating the user characteristics is, for example, information indicating the personality of the user related to the advice. Specifically, information such as “prefers strict advice” or “prefers gentle comments”. Information indicating user characteristics is also used for advice.

  Prior to the comment to be advised, the user information input unit 12 inputs a personality name that is information indicating the personality of the user. For example, the user information input unit 12 determines the personality of the user based on a questionnaire when the user is registered in the server 10 and inputs the personality name. Alternatively, the user information input unit 12 may receive the personality name transmitted from the client terminal 20 and input the personality name. The user information input unit 12 stores the input personality name in the storage unit 11 in association with the user ID.

  The advice candidate generation unit 13 is advice candidate generation means for generating advice candidates for the user based on the user information input by the user information input unit 12. Advice candidates are generated for each of the three advice categories as follows.

  Candidates for advice (advice by dish) according to the food that the user ate in the first advice category are generated as follows. The advice candidate generation unit 13 extracts information indicating food from the comment input by the user information input unit 12, and generates an advice candidate based on the extracted information. Specifically, the advice candidate generation unit 13 reads from the storage unit 11 the advice waiting data input by the user information input unit 12 and stored in the storage unit 11.

  The advice candidate generation unit 13 extracts a dish name that is a description indicating a dish (a type of food) from a comment included in the advice waiting data. For example, the advice candidate generation unit 13 performs morphological analysis on the comment. If the comment is "I ate ramen and dumplings today", the morpheme "today", "ha", "ramen", "to", "gyoza", "wo", "eating", "wasted" can get. The obtained morpheme is compared with the dish name, and if the morpheme matches the dish name, the dish name is extracted. In the above example, “ramen” and “gyoza” are extracted as cooking names. The dish name is stored in the storage unit 11 in advance. For example, the information of the dish master shown in FIG. 2B is stored for each dish.

  As shown in FIG. 2B, the dish master is information in which a dish ID, a dish name, a dish category, a nutritional component, and a priority order are associated with each other. The dish ID is information for identifying the dish. The dish name is the name of the dish. The dish category is information indicating the classification of the dish. For example, for a dish with “baked rice balls”, the cooking category is “rice balls”. The nutritional component is a value indicating the amount of each nutritional component of the dish. Nutritional components are, for example, carbohydrates, proteins, lipids and the like. The priority order is information indicating the order of the dishes among the dishes. For example, a higher priority is set for a dish for which advice is effective. When “Ramen” and “Rice” are eaten, the advice for “Ramen” is more effective than “Rice”, so the priority of “Ramen” is set higher. Information on the cooking master is stored in the storage unit 11 in advance by an administrator of the server 10 or the like.

  The advice candidate generation unit 13 reads the dish master from the storage unit 11 and uses it for extracting the dish name. The morphological analysis can use a conventional method. In the morpheme analysis, it is also possible to perform unique customization so that a dish name is obtained as one morpheme.

  In the storage unit 11, advice candidates are stored in advance for each dish name. For example, information on the dish advice data shown in FIG. 2C is stored for each dish name. As shown in FIG. 2C, the dish advice data is information in which dish names and advice candidates are associated with each other. The advice candidate is text that is a candidate for advice. The cooking advice data is stored in the server 10 in advance by an administrator of the server 10 or the like. However, the cooking advice data can be changed as will be described later.

  The advice candidate generation unit 13 reads out advice candidates stored in the storage unit 11 in association with the extracted dish name as dish advice data. When a plurality of dish names are extracted, an advice candidate associated with the dish name having the highest priority is read out. In the above example, “Ramen” and “Gyoza” are extracted as the names of the dishes, and among these, the advice candidate “higher priority” associated with “Ramen” should be eaten. ” Is read out. A plurality (for example, four) of advice candidates may be associated with one dish name. In that case, the advice candidate generation unit 13 reads one advice candidate at random from a plurality of advice candidates. The read advice candidates are cooking-specific advice candidates. If a dish name is not extracted from the comment, no advice candidate is generated. Alternatively, in this case, general advice candidates may be generated regardless of the dish (the same applies to the second advice category below).

  A candidate of advice (nutrient-based advice) according to the nutritional component taken by the user as a second advice category is generated as follows. The advice candidate generation unit 13 calculates an amount for each nutrient component based on the extracted information, and generates an advice candidate based on the calculated amount. The advice candidate generation unit 13 reads a value indicating the amount of each nutritional component for the extracted dish with the dish name from the dish master of the storage unit 11. When a plurality of dish names are extracted, the advice candidate generating unit 13 takes the sum of these values for each nutrient component. In the above example, the sum of the value of the nutritional component of “ramen” and the value of the nutritional component of “dumpling” is taken. This value is considered to be the amount of nutritional component that the user has consumed in one meal.

  In the storage unit 11, a value indicating the amount of the intake standard for each nutritional component is stored in advance by the administrator of the server 10 (not shown). The advice candidate generation unit 13 reads the value from the storage unit 11 and compares the value for each nutritional component based on the comment with the value of the intake standard. Note that the value of the intake standard may be a value according to the sex, age, etc. of the user. The advice candidate generation unit 13 selects a nutritional component having a characteristic nutritional component value based on the comment as a nutritional component used for generating a candidate for advice. For example, the advice candidate generator 13 advises the nutritional component whose values are most dissimilar (for example, the ratio (%) of the value based on the comment with respect to the intake standard is the farthest from 1 (100%)). It shall be used for generating candidates.

  In the storage unit 11, advice candidates are stored in advance for each nutritional component. For example, information on nutritional component advice data shown in FIG. 2D is stored for each nutritional component. As shown in FIG. 2D, the nutritional component advice data is information in which nutritional component names and advice candidates are associated with each other. The advice candidate is text that is a candidate for advice. The nutritional component advice data is stored in the server 10 in advance by an administrator of the server 10 or the like. However, the nutritional component advice data can be changed as will be described later. The advice candidates of the nutritional component advice data may be stored for each case where the nutritional component is excessive or insufficient. Further, advice candidates may be provided for each degree of excess or deficiency.

  The advice candidate generation unit 13 reads out the advice candidates stored in the storage unit 11 in association with the specified nutritional component as the nutritional component advice data. In the above example, the advice candidate “Let's be careful not to eat too much carbohydrate” is read for “Carbohydrate”. Regarding advice candidates based on nutritional components, one advice candidate may be randomly read from a plurality of advice candidates as in the case of a dish name. The read advice candidate is a candidate for nutritional component advice.

  Candidates for advice (personal advice) according to the personality of the user, which is the third advice category, are generated as follows. The advice candidate generating unit 13 reads the personality name stored in the storage unit 11 in association with the user ID included in the advice waiting data.

  In the storage unit 11, advice candidates are stored in advance for each character of the user. For example, information on personality advice data shown in FIG. 2E is stored for each personality of the user. As shown in FIG. 2E, the personality advice data is information in which personality names and advice candidates are associated with each other. The advice candidate is text that is a candidate for advice. The personality advice data is stored in the server 10 in advance by an administrator of the server 10 or the like. However, as will be described later, the information of the personality advice data can be changed.

  The advice candidate generation unit 13 reads out advice candidates stored in the storage unit 11 in association with the read personality name as information of personality advice data. In the above example, the advice candidate “Let's do our best from tomorrow!” Is read for the personality “I like strict advice”. As for advice candidates based on personality, one advice candidate may be read at random from a plurality of advice candidates as in the case of a dish name. The read advice candidates are personality advice candidates.

  The advice candidates of the respective advice categories generated as described above are connected to become advice candidates for comments included in the advice waiting data. In the above example, “You should eat vegetables. Be careful not to eat too much carbs. Let's do our best tomorrow!” Is a candidate for advice. Note that the line where the advice category candidates are connected may include a line feed. The advice candidate generation unit 13 stores the generated advice candidates in the storage unit 11 as advice candidate data illustrated in FIG.

  As shown in FIG. 2F, the advice candidate data is information in which a comment ID, a candidate ID, an advice category, and a candidate sentence are associated with each other. The comment ID is a comment ID of a comment that is a target of advice. The candidate ID is information for specifying a candidate for advice, and is given, for example, when the advice candidate generating unit 13 stores advice candidate data in the storage unit 11. The advice category is one of the above three advice categories. Further, the advice classification of the advice candidate data may include the detailed classification used for generating the advice candidates, and in the above example, “ramen”, “carbohydrate”, and “prefer strict advice”. The candidate sentence is an advice candidate text generated by the advice candidate generator 13.

  The advice candidates generated by the advice candidate generation unit 13 may be in accordance with previous advice from a supporter and a response from the user. Specifically, it will be described later.

  The advice candidate output unit 14 is an advice candidate output unit that outputs the advice candidates generated by the advice candidate generation unit 13 to the supporter. The advice candidate output unit 14 outputs an advice candidate by transmitting the advice candidate to the client terminal 20 of the supporter. The advice candidate is transmitted to the client terminal 20 in accordance with the operation of the supporter to the client terminal 20.

  For example, when the client terminal 20 of the supporter accesses the server 10, the supporter top screen shown in FIG. 4A is displayed on the client terminal 20. A list of users waiting for advice is displayed on the supporter top screen. This display is performed based on advice waiting data, for example. In this list, a user name, information about the user (age, occupation, height, weight, etc.), a comment, and the like are displayed. This list may be, for example, new arrival order, posting order, unanswered limitation, random, and the like. When any user (comment) is selected by a supporter operation on the supporter top screen, a screen for the selected user shown in FIG. 4B is displayed. In this screen, when the button B3 that is displayed as “advice” is pressed by a supporter operation, the comment ID of the displayed comment is transmitted from the client terminal 20 to the server 10.

  The advice candidate output unit 14 receives the comment ID. The advice candidate output unit 14 reads advice candidate data including the received comment ID from the storage unit 11. The advice candidate output unit 14 generates information shown in FIG. 5A for displaying the candidate sentence of the read advice candidate data as an advice candidate, and transmits the information to the client terminal 20 of the supporter. In the client terminal 20, the information is received and displayed. On the screen on which this information is displayed, a form F2 for displaying advice candidates as default values and inputting advice is provided.

  The advice input unit 15 is an advice input unit that inputs advice corresponding to the advice candidate output by the advice candidate output unit 14 from the supporter. The advice input unit 15 inputs advice that has been corrected for the advice candidates. The advice input unit 15 receives and inputs the advice transmitted from the client terminal 20 of the supporter. The advice is transmitted from the client terminal 20 in accordance with the operation of the supporter to the client terminal 20.

  As described above, the client terminal 20 of the supporter displays a screen shown in FIG. 5A on which a candidate F and a form F2 for inputting advice are displayed. The advice candidates displayed on the form F2 can be rewritten by the operation of the supporter (not necessarily rewritten). On this screen, when the button B4 displayed as “Post” is pressed by a supporter operation, the text described in the form F2 is transmitted from the client terminal 20 to the server 10. When advice is sent from the client terminal 20 to the server 10, it is sent together with the user ID of the supporter.

  The advice input unit 15 stores the received advice in the storage unit 11 as transmission advice data illustrated in FIG. As illustrated in FIG. 2G, the transmission advice data is information in which an advice ID, a comment ID, an advice category, an advice sentence, and a correction presence / absence flag are associated with each other. The advice ID is information for specifying the advice input from the supporter, and is given, for example, when the advice input unit 15 stores the transmission advice data in the storage unit 11. The comment ID is the comment ID of the comment of the user who is the target of advice input from the supporter. The advice category and the advice sentence are the advice category of the text included in the advice input from the supporter and the text itself. Thus, the advice input from the supporter is stored for each advice category in the transmission advice data. The advice classification is determined by the advice input unit 15 as shown below. The correction presence / absence flag is information indicating whether or not the advice has been corrected from the advice candidates. Further, in the server 10, the advice input from which supporter can be grasped by the user ID of the supporter.

  The advice input unit 15 determines an advice category for the text that constitutes the input advice as follows. That is, the advice input unit 15 determines whether the text corresponds to (1) advice by dish, (2) advice by nutritional component, (3) advice by character, or none.

  In the storage unit 11, a delimiter character for delimiting a sentence is stored in advance by an administrator of the server 10 or the like (not shown). The delimiter characters are, for example, line feeds, “.”, “!”, Pictograms, and emoticons with text such as “(^^)”. The advice input unit 15 reads a delimiter from the storage unit 11 and divides the input advice into a plurality of texts using the delimiter. Note that consecutive delimiters are regarded as one delimiter. In addition, the delimiter character constitutes the last part of the text before and after the delimiter.

  The advice input unit 15 reads a delimiter character from the storage unit 11. First, the advice input unit 15 determines whether or not the number of line breaks included in the advice is equal to 2 of the number of advice classifications−1 (3-1). That is, the advice input unit 15 determines whether or not the advice is divided into three texts by a line feed. If the advice input unit 15 determines that the number of line breaks included in the advice is 2, the advice input unit 15 selects (1) advice for each dish, (2) advice for each nutritional ingredient, in order from the front of the text in which the advice is separated by line breaks, ( 3) Judge as personality advice.

  When it is determined that the number of line breaks included in the advice is not 2, the advice input unit 15 determines whether or not the advice is divided into three texts by a delimiter including a line break. If the advice input unit 15 determines that the advice is divided into three texts, the advice input unit 15 sequentially (1) advice by cooking, (2) advice by nutritional ingredient, and (3) advice by personality from the text in which the advice is divided. It is judged that.

  When it is determined that the advice is not divided into three texts, the advice input unit 15 determines whether or not each of the divided texts matches the advice candidate of each advice category. The advice input unit 15 determines that the text corresponding to the advice candidate is in the advice category of the advice candidate.

  For texts for which the advice category could not be determined as described above, a determination is made based on keywords according to the advice category as follows. In the storage unit 11, information in which a keyword (word) and a score are associated with each advice category is stored in advance by an administrator of the server 10 or the like. This keyword is expected to appear in the text in each advice category. Keywords that appear to have a higher degree of appearance are given higher scores. For example, (1) For the cooking another advice, because that is provided to the top of the advice, keywords related to greeting (for example, "Hello") and higher scores to. In addition, (2) the advice for each nutritional component is provided at an intermediate position of the advice, so the score is set so that it is not classified into this category as much as possible.

  The advice input unit 15 reads the above information from the storage unit 11. The advice input unit 15 determines whether or not each keyword is included in the text. If the keyword is included, the score of the keyword is set as the score of the text. The score is calculated for each advice category. When a plurality of keywords are included, the score of the text is the sum of them.

  The advice input unit 15 determines that the advice category having the highest score for each text is the advice category of the text. If the same text has the same score for multiple advice categories, the advice categories are in the following order: (1) Advice by cuisine, (3) Personality advice, and (2) Nutrition component advice. Distribute. When a plurality of texts are determined to be the same advice category, if they are continuous, they are combined. If they are not continuous, the first half text is set as the advice category, and the second half text is set as no advice category. The advice input unit 15 sets the text that cannot be determined as described above to no advice category. In addition, when the advice is a single text, that is, when the advice is not divided, the advice input unit 15 may determine that the text is (1) advice by dish.

  For example, if the advice that was entered was "You may eat a little too much. <New line> Take care to eat too much carbs. <New line> Let's work hard tomorrow!" You may eat too much. <Line break> "Be careful not to eat too much carbs. <Line break>" "Let's do our best from tomorrow!" Judgment according to cooking, (2) advice according to nutritional ingredients, and (3) advice according to personality.

  In addition, entered advice, "Hello! <New line> It might be too much to eat a little bit. The <new line> take too much of carbohydrates a little be careful. <New line> Let's do our best!" Was that case, "Hello! <new line>" (sentence 1) "it might be too much to eat a little. <new line>" (sentence 2) "is to take too much of carbohydrates will help you a little attention. <new line>" (sentence 3 ) “Let's do our best!” (Sentence 4) is divided into four texts, Sentence 1 to Sentence 4. In this case, the score for each advice category is calculated for each text. For example, sentence 1 = (1) 5 dishes, (2) 0 nutrition, (3) 0 personality, sentence 2 = (1) 3 dishes, (2) 1 nutrition, (3) 0 personality, Sentence 3 = (1) 1 dish, (2) 5 nutrition, (3) 0 personality, 4 = (1) 0 cooking, (2) 0 nutrition, (3) 5 personality, etc. A score is calculated. For each text, the advice category with the highest score is determined. Sentence 1 and sentence 2 are (1) advice by cuisine, sentence 3 is (2) nutritional ingredient advice, and sentence 4 is (3) personality advice.

  The advice input unit 15 reads the advice candidate data from the storage unit 11, and determines whether or not each text has been changed from the advice candidates for each advice category. The advice input unit 15 stores the transmission advice data illustrated in FIG. 2G in the storage unit 11 based on the above determination. As for the advice classification of the transmission advice data, in addition to the above three classifications, more detailed classification information corresponding to advice candidates is stored. In the case of (1) advice by dish, information on the name of the dish corresponding to the advice candidate is stored, (2) in the case of advice by nutritional ingredient, in the name of a nutrient, and (3) in the case of personality advice, the information of personality name is stored. These pieces of information are obtained from detailed classifications used for generating advice candidates related to advice.

  The advice output unit 16 is an advice output unit that outputs the advice input by the advice input unit 15 to the user. The advice output unit 16 outputs the advice by transmitting the advice to the client terminal 20 of the user. The transmission of advice to the client terminal 20 is performed in response to a user operation on the client terminal 20.

  For example, when the user's client terminal 20 accesses the server 10, the advice output unit 16 reads the transmission advice data shown in FIG. 2G for the user's comment (including the comment ID of the comment) from the storage unit 11. . The advice output unit 16 generates information shown in FIG. 5B for displaying the advice sentence of the read transmission advice data, and transmits the information to the client terminal 20 of the user. In the client terminal 20, the information is received and displayed.

  The response input unit 17 is a response input unit that inputs a response from the user according to the advice output by the advice output unit 16. The response input unit 17 receives and inputs a response transmitted from the user's client terminal 20. The response from the client terminal 20 is transmitted in response to a user operation on the client terminal 20.

  A button B5 displaying “thank you” is provided on the screen shown in FIG. This button B5 is a button for transmitting a response from the user to the server 10. The button B5 is pressed when the user expresses a gratitude to the supporter who gave the advice, for example, the user thinks that the advice is a good advice. For example, points are given to supporters who have given advice when the response is received. By accumulating the points, the supporter can receive some benefit, for example, goods or services from the administrator of the server 10 or receive a discount on them. In this way, a reward can be given to a supporter who has given good advice to provide an incentive for giving advice.

  The response input unit 17 updates the data stored in the storage unit 11 based on the reception of the response. The response input unit 17 refers to the transmission advice data shown in FIG. 2G stored in the storage unit 11 for the advice text relating to the received response, and refers to the information of the correction presence / absence flag for each advice category. Then, it is determined whether the advice candidate has been changed. If the response input unit 17 determines that the advice text has been changed from the advice candidate, the response input unit 17 determines whether the storage unit 11 stores the advice category and the learning data for the text.

  The learning data is information used for updating advice candidates generated thereafter. FIG. 2 (h) shows the learning data. As shown in FIG. 2H, the learning data is information in which candidate IDs, advice categories, correction sentences, and reward counts are associated. The candidate ID is an ID for specifying a record of learning data, and is given, for example, when the response input unit 17 stores the learning data in the storage unit 11. The advice category and the corrected sentence are information corresponding to the advice category and the advice sentence in FIG. The reward count is the number of times a response “thank you” has been made to the advice including the revised sentence.

  FIG. 6A shows a specific example of learning data. The leftmost column is the advice category. In this example, the advice category is (1) a category in which the dish name of the advice for each dish is “ramen”. The next column is the revised text. The rightmost column is the reward count.

  When the response input unit 17 determines that the advice text has been changed from the advice candidates, the storage unit 11 stores the learning data in which the advice category and the text are the advice category and the corrected sentence. It is judged whether it is done. When the learning data is stored in the storage unit 11, the response input unit 17 increases the reward count of the learning data stored in the storage unit 11 by one. The reward count may be increased by 1 even when the text and the corrected sentence do not completely match. For example, in consideration of the case where the user's name is added to the advice, the reward count may be increased by 1 when the entire corrected sentence matches a part of the text.

  When the learning data is not stored in the storage unit 11, the response input unit 17 causes the storage unit 11 to newly store learning data in which the advice category and the text are the advice category and the corrected sentence. The reward count of the learning data is 1.

  When the response input unit 17 determines that the advice text has not been changed from the advice candidate, the response input unit 17 updates the advice classification and the transmission advice candidate data for the text stored in the storage unit 11.

  Similarly to the learning data, the transmission advice candidate data is information used for updating the advice candidates generated thereafter. FIG. 2 (i) shows the transmission advice candidate data. As shown in FIG. 2I, the transmission advice candidate data is information in which a candidate ID, an advice category, a candidate sentence, and a reward count are associated with each other. The candidate ID is an ID for specifying a record of transmission advice candidate data, and is set in advance. The advice category and the candidate sentence are information corresponding to the advice category and the advice sentence in FIG. The reward count is the number of times a response “thank you” is made to the advice including the candidate sentence. Note that the candidate sentence matches one of the advice candidates of the dish advice data, nutritional component advice data, and personality advice data in FIGS. 2 (c) to 2 (e). When an advice candidate is newly set, transmission advice candidate data for the advice candidate is newly generated. The reward count of newly generated transmission advice candidate data is set to zero.

  FIG. 6B shows a specific example of transmission advice candidate data. The leftmost column is the advice category. In this example, the advice category is (1) a category in which the dish name of the advice for each dish is “ramen”. The next column is a candidate sentence. The rightmost column is the reward count.

  The response input unit 17 is a reward count of the advice advice data stored in the storage unit 11 in which the text of the advice related to the received response and the advice classification of the text are candidate sentences and advice classification Increase by one.

  The advice candidate generation unit 13 generates the advice candidate based on the advice input by the advice input unit 15 before the generation of the advice candidate. The advice candidate generation unit 13 generates the advice candidate based on the response input by the response input unit 17 before the generation of the advice candidate.

  Specifically, the advice candidate generation unit 13 is based on the learning data and the transmission advice candidate data in which the correction and response from the advice candidate are reflected as described above, and FIG. Update the food advice data, nutritional advice data, and personality advice data. The advice candidate generation unit 13 performs the following processing, for example, every preset time (for example, every several hours).

  The advice candidate generation unit 13 identifies learning data having a reward count of 2 or more from the learning data stored in the storage unit 11. The corrected sentence of the learning data is set as a new advice candidate. The learning data having a reward count of 2 or more is advice text that is input (corrected) by the supporter multiple times by the supporter and responded by the user. Therefore, since it may be used as advice in the future, it will be a candidate for advice in the future. In addition, 2 which is said threshold value may be another value (for example, 3 or more values). Also, other specific numerical values in the present embodiment may be appropriately changed according to the aspect.

  The advice candidate generation unit 13 specifies transmission advice candidate data having the same advice category as the advice category of the learning data from the transmission advice candidate data stored in the storage unit 11. The advice candidate generation unit 13 identifies transmission advice candidate data having the smallest reward count value among the transmission advice candidate data stored in the identified storage unit 11. The transmission advice candidate data becomes an advice candidate to be replaced. When there are a plurality of pieces of learning data having a reward count of 2 or more, transmission advice candidate data corresponding to that number is specified from the smaller reward count value.

  The advice candidate generation unit 13 deletes the specified learning data and transmission advice candidate data. Further, transmission advice candidate data is newly generated with the advice classification and the corrected sentence of the specified learning data as the advice classification and the candidate sentence. The reward count of the transmission advice candidate data may be 0, or the reward count value of the learning data may be taken over.

  Further, the advice candidate generation unit 13 receives the cooking advice data, nutritional component advice data, or personality advice data of FIGS. 2C to 2E corresponding to the advice classification and candidate sentence of the specified transmission advice candidate data. delete. Moreover, the advice candidate production | generation part 13 newly adds the dish advice data of FIG.2 (c)-FIG.2 (e) corresponding to the advice classification and correction sentence of the specified learning data, nutrition component advice data, or personality advice data. Generate. That is, the advice candidate generating unit 13 generates the above data in which the advice candidate in the advice category is a corrected sentence. Thereby, the advice candidate generating unit 13 can use the text that has been the above-described corrected sentence when generating the candidate for advice in the future.

  In the case of the data shown in FIG. 6, advice candidates for the “ramen” category are replaced. As shown in FIG. 6A, “You may eat a little too much” is the text included in the advice that was input (corrected) by the supporter twice and responded by the user. Also, as shown in FIG. 6B, “Do not eat!” Is a text that has the smallest number of responses (one time) among the current advice candidates. Therefore, for this category, “You may eat a little too much” is replaced with “Do not eat!” And become a candidate for advice. The above is the function of the server 10 according to the present embodiment.

  Subsequently, an advice support method that is a process executed by the server 10 according to the present embodiment (an operation method of the server 10) will be described with reference to flowcharts of FIGS. First, processing when a comment is input from the user will be described with reference to the flowchart of FIG.

  When a comment is transmitted from the user's client terminal 20, the comment is received and input by the user information input unit 12 (S01, user information input step). Subsequently, the advice candidate generation unit 13 performs morphological analysis of the comment (S02, advice candidate generation step). Subsequently, the advice candidate generation unit 13 generates a dish-specific advice candidate based on the result of the morphological analysis (S03, advice candidate generation step). Moreover, the advice candidate production | generation part 13 produces | generates the candidate of advice according to nutrient composition based on the result of a morphological analysis (S04, advice candidate production | generation step). The advice candidate generation unit 13 generates personality-specific advice candidates based on the user's personality (S05, advice candidate generation step). The generated advice candidates are stored in the storage unit 11 (S06). The above is the processing when a comment is input from the user.

  Next, a process when advice candidates are output to the supporter and advice is input from the supporter will be described with reference to the flowchart of FIG. When the server 10 is accessed from the client terminal 20 of the supporter and the comment ID to be advised is transmitted from the client terminal 20 to the server 10, the comment candidate output unit 14 receives the comment ID. Subsequently, the advice candidate output unit 14 reads out the advice candidate corresponding to the received comment ID from the storage unit 11 and transmits it to the client terminal 20 of the supporter (S11, advice candidate output step).

  When advice is transmitted from the client terminal 20 of the supporter in response to the transmission of the advice candidate, the advice is received and input by the advice input unit 15 (S12, advice input step). The received advice is stored in the storage unit 11 after a change from the advice category and advice candidates is determined (S13). The above is the processing when advice candidates are output to the supporter and advice is input from the supporter.

  Next, processing when advice is output to the user will be described using the flowchart of FIG. When the server 10 is accessed from the user's client terminal 20, the advice output unit 16 reads out advice candidates for the user's comments from the storage unit 11 and transmits them to the user's client terminal 20 (S21, Advice output step). When a response of “thank you” is transmitted from the client terminal 20 of the user in response to the transmission of the advice candidate, the response input unit 17 receives the response (YES in S22, response input step). In this case, subsequently, the response input unit 17 updates the learning data or the transmission advice candidate data for each advice category in the storage unit 11 (S23). When a response “thank you” is not transmitted from the client terminal 20 of the user in response to the transmission of the advice candidate, the above processing is not performed. The above is the processing when advice is output to the user.

  Next, processing when an advice candidate is updated will be described using the flowchart of FIG. When the preset time is reached, the advice candidate generation unit 13 reads the learning data and the transmission advice candidate data stored in the storage unit 11, and replaces the advice candidates based on them (S31, S31). Advice candidate generation step). The above is the processing when the advice candidate is updated.

  As described above, in the present embodiment, advice candidates can be presented to the supporter. Thereby, a general user who is not an expert such as a registered dietitian can be a supporter. In addition, by presenting the advice candidates to the supporter, it is possible to reduce the burden for providing the supporter's advice.

  According to the present embodiment, the problem that the number of advisors is insufficient is solved for a user who wants to seek advice, consult, or encourage a healthy life. In addition, people who are interested in health and want to help someone can participate as supporters. In addition, since advice is finally transmitted from the supporter, humane communication can be secured. Thereby, advice between users can be activated. That is, according to the present embodiment, it is possible to provide advice so that many users can maintain motivation in health care and the like.

  Moreover, it is good also as extracting the cooking name which is the information regarding food from a comment like this embodiment, and producing | generating the advice candidate. Furthermore, it is good also as producing | generating the candidate of advice based on the quantity for every nutrient component. Moreover, it is good also as producing | generating the candidate of advice based on the information which shows user characteristics, such as a user's character. According to these configurations, it is possible to present appropriate advice candidates to the supporter, and appropriate advice is provided between users. However, the generation of advice candidates is not limited to that described above, and any method may be used as long as it is performed based on user information about the user. As described above, in the present embodiment, advice candidates are generated in three advice categories: (1) advice by dish, (2) advice by nutritional ingredient, and (3) advice by personality. It is good also as generating the candidate of advice in. Further, the number of advice categories is not necessarily three as in the present embodiment, and may be increased to four or five.

  Further, as described above, advice candidates may be generated based on the correction of the comment from the supporter. According to this configuration, the advice candidate can be corrected to the supporter, and an appropriate advice candidate can be presented according to the corrected advice.

  Further, advice candidates may be generated based on a response from the user. According to this configuration, it is possible to present an appropriate advice candidate according to the response from the user, that is, the user's reaction such as whether the advice is useful.

  By updating the advice candidates in a learning manner in this manner, appropriate advice candidates can be continuously generated. However, a configuration may be adopted in which the advice candidates are not updated.

  In the above-described embodiment, the function related to the advice has been described as the function of the server 10, but the server 10 may have a function related to health management other than the function related to advice. The advice support system according to the present invention may relate to advice in a field other than the above-described health management.

  DESCRIPTION OF SYMBOLS 10 ... Server, 11 ... Memory | storage part, 12 ... User information input part, 13 ... Advice candidate production | generation part, 14 ... Advice candidate output part, 15 ... Advice input part, 16 ... Advice output part, 17 ... Response input part, 20 ... Client terminal.

Claims (7)

User information input means for inputting user information about himself / herself from the advised user who is the user who receives advice,
Advice candidate generating means for generating candidate advice for the advised user based on the user information input by the user information input means;
Advice candidate output means for outputting the advice candidates generated by the advice candidate generation means to an advice user who is an advice user;
Advice input means for inputting advice according to advice candidates output by the advice candidate output means from the advice user;
Advice output means for outputting the advice input by the advice input means to the advised user; and
Advice support system with The user information input means inputs text as user information,
The advice support system according to claim 1, wherein the advice candidate generation unit extracts information indicating food from the text input by the user information input unit, and generates an advice candidate based on the extracted information.   The advice support system according to claim 2, wherein the advice candidate generation unit calculates an amount for each nutrient component based on the extracted information, and generates an advice candidate based on the calculated amount.   The advice support system according to any one of claims 1 to 3, wherein the user information input unit inputs information indicating user characteristics as user information. The advice input means inputs advice that has been corrected for a candidate for advice,
5. The advice candidate generation unit generates the advice candidate based on the advice that has been input by the advice input unit before the generation of the advice candidate and has been corrected. 6. The advice support system described in one item. Response input means for inputting a response from the advised user according to the advice output by the advice output means,
6. The advice according to claim 1, wherein the advice candidate generation unit generates the advice candidate based on a response input by the response input unit before generation of the advice candidate. Support system. An advice support method that is an operation method of the advice support system,
A user information input step for inputting user information about the user from the advised user who is the user who receives the advice;
Based on the user information input in the user information input step, an advice candidate generation step for generating advice candidates for the advised user;
An advice candidate output step for outputting the advice candidates generated in the advice candidate generation step to an advice user who is a user who performs advice;
An advice input step for inputting advice from the advice user according to the advice candidate output in the advice candidate output step;
An advice output step for outputting the advice input in the advice input step to the advised user; and
Advice support method including.

Images