JP7208457B2 - Information processing system, information processing method, information processing program, and information processing apparatus - Google Patents

Description

The present invention relates to an information processing system, an information processing method, an information processing program, and an information processing apparatus.

Gambling addiction has become a social problem, and there is a possibility that the number of patients with gambling addiction will increase if casinos are realized in Japan in the future. In recent years, addiction to online games and the Internet has also become a major social problem. Hereinafter, these behavioral addictions will be collectively referred to as addictions such as gambling.

Gambling addiction, like alcoholism and drug addiction, needs to take measures according to its severity. For example, for healthy and mildly at-risk individuals, guidance through educational institutions and counseling is effective. Counseling and treatment at a medical institution are effective for moderate-risk and severe-risk individuals. Furthermore, for severely at-risk individuals and addicts, it is effective to continuously participate in recovery facilities and self-help groups in addition to hospitals and other medical institutions.

Conventionally, there is known a technique in which a gambling machine or the like monitors an identified person's behavior such as gambling and determines the category of the monitored person's behavior such as gambling (see, for example, Patent Document 1).

Japanese Patent Publication No. 2007-503631

However, in Patent Literature 1 and the like, no consideration is given to displaying information regarding the degree of dependence on gambling or the like in an easy-to-understand manner.

In one aspect, an object of the present invention is to provide an information processing system, an information processing method, an information processing program, and an information processing apparatus capable of displaying information regarding the degree of dependence on a certain object in an easy-to-understand manner.

In one aspect, an information processing system includes an acquisition unit that acquires an index indicating how much information on behavior for each of a plurality of items related to a person differs from a standard , each of the plurality of items obtained by the obtaining unit , the degree of contribution indicating the degree of contribution of each of the plurality of items to the magnitude of the value indicating the degree of dependence of the certain person on the certain object obtained from the information; a calculation unit that calculates by multiplying the index of each of the plurality of items by weight information of each of the plurality of items stored in a storage unit; and a display control unit for displaying some or all of the items included in the .

Information about the degree of dependence on a certain target can be displayed in an easy-to-understand manner.

It is a figure showing roughly composition of an information processing system concerning one embodiment. 2 is a diagram showing the hardware configuration of a server owned by the ICT infrastructure operation organization of FIG. 1; FIG. FIG. 2 is a functional block diagram of a server owned by the ICT infrastructure operation organization of FIG. 1; 4 is a flowchart showing processing of an explanatory variable creation unit, an objective variable creation unit, and a prediction model creation unit; FIG. 5(a) is a diagram showing entrance/exit information to a casino facility, FIG. 5(b) is a diagram showing gaming information, and FIG. 5(c) is a diagram showing wide area action information. FIG. 6(a) is a diagram showing monthly explanatory variables (entrance/exit information) for each individual, and FIG. 6(b) is a diagram showing monthly explanatory variables (gaming information) for each individual. 6(c) is a diagram showing monthly explanatory variables (wide-area behavior information) for each individual. FIG. 7(a) is a diagram showing demographic information, and FIG. 7(b) is a diagram showing monthly explanatory variables (demographic information) for each individual. It is a figure which shows the monthly explanatory variable (merge version) for every individual. FIG. 9(a) is a diagram showing the person/family application information, FIG. 9(b) is a diagram showing visit restriction information, and FIG. 9(c) is a diagram showing event information. 9(d) is a diagram showing monthly target variables for each individual. It is a figure which shows the data for learning. It is a figure which shows an importance table. It is a flowchart which shows the process of a prediction model application part. It is a figure which shows an example of the explanatory variable created from the newly obtained data. FIG. 14(a) is a diagram showing prediction results, and FIG. 14(b) is a diagram showing an explanatory variable type table. 10 is a flowchart showing processing (part 1) of an output unit; FIG. 16 is a flowchart showing specific processing in step S120 of FIG. 15; FIG. FIG. 17(a) is a diagram showing a z-score table created based on the explanatory variables in FIG. 13, and FIG. 17(b) is a diagram showing a contribution degree table. FIG. 11 is a diagram (part 1) showing a list screen; FIG. 11 is a diagram (part 2) showing a list screen; FIG. 16 is a flow chart showing specific processing in step S130 of FIG. 15; FIG. FIG. 11 is a diagram (part 1) showing a detailed screen; FIG. 11 is a diagram (part 2) showing a detailed screen; 10 is a flowchart showing processing (2) of the output unit; FIG. 24(a) is a diagram showing an emphasizing method table according to risk score, and FIG. 24(b) is a diagram showing an emphasizing method table according to risk. FIG. 24 is a diagram showing an example of a screen displayed in step S208 of FIG. 23; FIG. FIG. 26 is a diagram showing a modification of the screen of FIG. 25;

An embodiment of an information processing system will be described in detail below with reference to FIGS. 1 to 26. FIG.

FIG. 1 schematically shows the configuration of an information processing system 100 according to one embodiment. As shown in FIG. 1, the information processing system 100 includes public institutions such as national and local governments, gambling addiction countermeasures, IR (Integrated Resort) operators, external institutions (medical institutions, counselors, external organizations). etc.), the ICT infrastructure operation organization gathers various information (big data) from the equipment installed in the commercial facility operator via the network N, analyzes it, and provides information on gambling addiction etc. .

As shown in FIG. 1, the IR operator has a server 20, a plurality of devices 22 connected to the server 20, and a terminal 24. The equipment 22 is an entry/exit gate provided at the entrance/exit of a casino facility operated by an IR operator, an electronic gaming machine such as a slot machine, a table game, and the like. The terminal 24 is a terminal used by IR operator staff during business. The server 20 transmits the information obtained from the device 22 to the server 80 owned by the ICT infrastructure operation organization. At the same time, the server 20 transmits the information obtained from the device 22 to the server 10 managed by the public institution and the server 30 managed by the gambling addiction prevention agency. In addition, the server 20 sends information applied to the casino facility by the user himself/herself and his/her family who use the casino facility, and information grasped by the staff in the casino facility input from the terminal 24 to the ICT-based server 80. Send. At the same time, the server 20 transmits the information to the server 10 managed by a public institution and the server 30 managed by a gambling addiction prevention agency. Here, the information obtained from the device 22 includes an entry/exit log that records entry/exit times for each user, and gaming information that records game play times, bets, winnings, etc. for each user. . In addition, the information applied to the casino facility by the user himself/herself or his/her family includes regulation information regarding the user (maximum number of usage per month, maximum usage amount per month, etc.). Furthermore, the information that the staff in the casino facility inputs from the terminal 24 includes information that the user has been restricted from visiting, information on the user's attitude and behavior to be careful (Responsible Gambling (RG) application information), etc. is included.

A public institution has a server 10 and a terminal 12 . The server 10 acquires the entry/exit logs transmitted from the servers 20 of the IR operators located in multiple locations throughout the country, and generates entry/exit information to the casino facility as shown in FIG. 5(a). The entrance/exit information includes visit information of each user (information on who entered the casino facility when and when it exited). The information input to the terminal 12 includes information indicating that visit restrictions have been imposed on the user, information on the attitude/behavior to which the user should be careful (Responsible Gambling (RG) applied information), and the like. The server 10 transmits the entrance/exit information and the information obtained from the terminal 12 to the ICT-based server 80 . At the same time, the server 10 transmits the entrance/exit information and the information obtained from the terminal 12 to the server 30 managed by the gambling addiction prevention agency.

The external institution has a server 40 and a terminal 42 . Information about patients with gambling addiction or the like is input to the terminal 42 from a doctor, counselor, or the like belonging to an external organization. The information about patients includes counseling results, diagnosis and treatment results, contents of recovery assistance support, and the like. The server 40 transmits the information obtained from the terminal 42 to the ICT-based server 80 . At the same time, the server 40 transmits the information obtained from the terminal 42 to the server 10 managed by a public institution and the server 30 managed by a gambling addiction prevention agency.

The gambling addiction countermeasure organization has a server 30 and a terminal 32 . In the terminal 32, application information from the resident or his/her family (application information for the individual/family as shown in FIG. 9(a)) and personal attribute information (demographic information as shown in FIG. 7(a)) are input. be done. In addition, information (visit restriction information in FIG. 9B) indicating that the gambling addiction countermeasure agency has performed early intervention work (implementation of visit restriction) for the residents is input to the terminal 32 . The terminal 32 transmits the input information to the server 30 . The server 30 transmits the information obtained from the terminal 32 to the ICT-based server 80 .

A commercial facility operator has a server 70 . The commercial facility operator collects, in the server 70, wide-area activity information (see FIG. 5(c)) obtained from stores such as supermarkets operated by the commercial facility operator. The server 70 transmits the collected information to the ICT-based server 80 .

The ICT infrastructure operating organization has a server 80 and a terminal 82 . The server 80 is an ICT platform, and collects and analyzes information from the servers owned by each of the organizations described above to generate and output a screen displaying information about addiction such as gambling.

FIG. 2 shows the hardware configuration of the ICT-based server 80 . As shown in FIG. 2, the server 80 includes a CPU (Central Processing Unit) 90, a ROM (Read Only Memory) 92, a RAM (Random Access Memory) 94, a storage unit (here, a HDD (Hard Disk Drive)) 96, a network It has an interface 97, a display section 93, an input section 95, a portable storage medium drive 99, and the like. Each component of the server 80 is connected to the bus 98 . In the server 80, the CPU 90 executes a program (including an information processing program) stored in the ROM 92 or the HDD 96, or a program (including an information processing program) read from the portable storage medium 91 by the portable storage medium drive 99. By doing so, the function of each part shown in FIG. 3 is realized. Note that the function of each unit in FIG. 3 may be implemented by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). Devices (servers and terminals) other than the server 80 included in the information processing system 100 in FIG. 1 also have the same hardware configuration as in FIG.

FIG. 3 shows a functional block diagram of the ICT-based server 80. As shown in FIG. As shown in FIG. 3, in the server 80, the CPU 90 executes a program to obtain an information acquisition unit 50, an explanatory variable creation unit 51, an objective variable creation unit 52, a prediction model creation unit 53, a prediction model application unit 54, It functions as an acquisition unit, a calculation unit, and an output unit 55 as a display control unit.

The information acquisition unit 50 acquires information from the servers 10, 20, 30, 40, and 70 via the network N. FIG. The information acquisition unit 50 transmits the acquired information to the explanatory variable creation unit 51, the objective variable creation unit 52, and the prediction model application unit 54.

The explanatory variable creating unit 51 uses the information acquired by the information acquiring unit 50 to create explanatory variables used for machine learning.

The objective variable creating unit 52 uses the information acquired by the information acquiring unit 50 to create an objective variable used for machine learning.

The predictive model creating unit 53 creates a predictive model using the explanatory variables created by the explanatory variable creating unit 51 and the objective variables created by the objective variable creating unit 52 . The prediction model creating unit 53 stores the created prediction model in the prediction model DB 60 . In addition, the predictive model creating unit 53 stores the importance of each explanatory variable as weight information calculated when creating the predictive model in an importance table 62 as a storage unit. In FIG. 3, the prediction model DB 60 is described as if the server 80 has it, but the prediction model DB 60 is held by a device other than the server 80 (for example, a data server that can communicate with the server 80). good too. Similarly, the importance level table 62 may also be held by a device other than the server 80 .

The predictive model application unit 54 creates an explanatory variable from the information newly acquired by the information acquisition unit 50, and applies the predictive model to the created explanatory variable, thereby predicting the risk of addiction such as gambling. The prediction model application unit 54 transmits prediction results and explanatory variables to the output unit 55 .

The output unit 55 generates a screen for displaying the prediction result of the prediction model application unit 54 and displays (outputs) it on the display unit 93 or outputs it to the terminal 82 to display it on the display unit of the terminal 82 . The output unit 55 uses the prediction results and explanatory variables of the prediction model application unit 54 and the importance table 62 created by the prediction model creation unit 53 when generating the screen.

(Regarding processing of explanatory variable creation unit 51, objective variable creation unit 52, and prediction model creation unit 53)
Next, the processing of the explanatory variable creation unit 51, the objective variable creation unit 52, and the prediction model creation unit 53 will be described in detail along the flowchart of FIG. 4 and with appropriate reference to other drawings.

The process of FIG. 4 is, for example, a process of generating (updating) a prediction model using the acquired information when the information acquiring unit 50 acquires a certain amount of information. A certain amount of information can be, for example, information obtained over a predetermined period of time (eg, 3 months, 6 months, etc.). That is, the processing of FIG. 4 is performed each time a predetermined period of time elapses. Alternatively, the user can specify any period or timing and generate (update) the prediction model based on it.

In the process of FIG. 4, first, in step S10, the explanatory variable creation unit 51 creates explanatory variables. Specifically, the explanatory variable creation unit 51 processes and tabulates personal behavior information and personal attribute information, and creates explanatory variables used for prediction model generation. The individual behavior information is time-series event information that records when and what each individual did. to create For example, if the fixed period is 6 months, the aggregation process is performed sequentially while shifting the aggregation period of 6 months by one month, such as January to June 2017 or February to July 2017. .

The personal behavior information includes, for example, entry/exit information to a casino facility as shown in FIG. 5(a), gaming information as shown in FIG. 5(b), wide area behavior information as shown in FIG. and so on. The personal attribute information includes demographic information and the like as shown in FIG. 7(a).

The entry/exit information to the casino facility shown in FIG. 5(a) includes information on the entry time and exit time of each user to the casino facility. The explanatory variable creation unit 51 creates a monthly explanatory variable (entrance/exit information) for each individual as shown in FIG. Monthly explanatory variables (entrance/exit information) for each individual include the number of days and total number of days each user visited casino facilities during a certain period (for example, January to June 2017 or February to July 2017). Includes dwell time, average dwell time per day, and maximum dwell time per day.

The gaming information in FIG. 5(b) includes information on the game machine used by each user, playing time, bet amount, and winning amount. The explanatory variable creation unit 51 creates a monthly explanatory variable (gaming information) for each individual as shown in FIG. 6B by aggregating the gaming information. The monthly explanatory variables (gaming information) for each individual include the total number of bets made by each user, the total bet amount, the total It includes win amounts, average daily bets, and maximum single bets.

The wide-area action information in FIG. 5C includes information on the store used by each user, the type of use (purchasing goods or eating and drinking), the date and time of use, and the amount of money. The explanatory variable creation unit 51 creates monthly explanatory variables (wide-area activity information) for each individual as shown in FIG. 6(c) by aggregating the wide-area activity information. Monthly explanatory variables (wide-area behavior information) include the total number of purchases, total purchase amount, total number of times of eating and drinking for each user in a certain period (for example, January to June 2017 or February to July 2017). , including the average amount of food and drink per meal. Of the monthly explanatory variables for each individual shown in Figures 6(a) to 6(c), the aggregated results for the period from January to June 2017 are the explanations used for forecasting as of July 2017. variable. In addition, the tabulated results for the period from February to July 2017 will be used as explanatory variables for forecasting as of August 2017.

The demographic information in FIG. 7(a) includes information that may change over time (married/unmarried, residence type, etc.) and static information that does not change over time (gender, year of birth, etc.) as attribute information of each user. month and day) are included. The explanatory variable creation unit 51 creates monthly explanatory variables (demographic information) for each individual as shown in FIG. 7B by processing the demographic information. The monthly explanatory variables (demographic information) for each individual in FIG. and age information. The flag information includes a male flag indicating whether the person is male, a married flag indicating whether the person is married, a detached house flag indicating whether the person is in a detached house, an apartment flag indicating whether the person is in an apartment, a rental flag indicating whether the person is renting, etc. is included. The flag information is created by converting demographic information into dummy variables (decomposing a multi-valued category value into a plurality of pieces of flag information). Age information is calculated from date of birth information. Of the monthly explanatory variables (demographic information) for each individual in FIG. 7(b), the information for the period from January to June 2017 is the explanatory variable used for prediction as of July 2017. . Also, the information for the period from February to July 2017 is an explanatory variable used for prediction as of August 2017. If the information changes within the period of six months, the final state information is used, but if the information does not change, the past information may be used as it is.

The explanatory variable creation unit 51 merges the monthly explanatory variables for each individual created as described above (FIGS. 6(a) to 6(c) and FIG. 7(b)) into monthly explanatory variables (merged version) for each individual. The explanatory variable creation unit 51 transmits the created monthly explanatory variables (merged version) for each individual to the prediction model creation unit 53 .

Returning to FIG. 4, in step S12, the objective variable creation unit 52 creates an objective variable. Here, in the risk prediction of addiction such as gambling, it is important how to define the addiction risk. This embodiment aims at early detection of addiction such as gambling. Therefore, rather than the onset of addiction, the risk state is considered to be the state in which an event that is a precursor to addiction, such as a person's report, a family's report, or visit refusal information, is occurring. Binary classification (prediction) of whether or not it will occur is performed.

Specifically, the objective variable creation unit 52 creates the Generate event information as shown. Based on the event information of FIG. 9(c), the objective variable creating unit 52 creates monthly objective variables for each individual as shown in FIG. 9(d).

Here, the person/family application information in FIG. 9(a) indicates that the person or family has reported to the casino facility or gambling addiction countermeasures organization that admission to the casino facility is restricted due to the risk of addiction. and includes information on the user ID, applicant, and date of application. In addition, the visit restriction information in FIG. 9(b) monitors behavior suspected of addiction, such as the number of visits exceeding a threshold, and imposes preventive measures such as restricting admission and limiting the number of visits to casino facilities and gambling facilities. It is information that shows that the addiction countermeasure organization has done it. The visit restriction information includes information on the type of visit restriction and information on the visit restriction date in association with the user ID. The event information in FIG. 9(c) is information obtained by merging the person/family application information in FIG. 9(a) and the visit restriction information in FIG. 9(b).

The monthly objective variable for each individual in FIG. 9(d) includes a user ID, a prediction implementation month, and a correct flag. In this embodiment, the objective variable creation unit 52 extracts the date of occurrence of an event from the personal regulation application information. Create a correct flag (objective variable) so that it becomes "example". For example, if the fixed period (“prediction period”) is two months, the correct flag when July 2017 is the prediction implementation month is the event occurring in August or September 2017. If so, it is a positive example (correct flag=1), otherwise it is a negative example (correct flag=0). Also, the correct flag when August 2017 is the predicted implementation month is a positive example (correct flag = 1) if an event occurs in September or October 2017, otherwise is a negative example (correct flag=0).

If the purpose is to predict the occurrence of the first event, neither the explanatory variable nor the objective variable after the occurrence of the event shall be used.

Returning to FIG. 4, in the next step S14, the prediction model creating unit 53 creates learning data. In this case, the predictive model creating unit 53 creates learning data by combining columns of the explanatory variables created by the explanatory variable creating unit 51 and the objective variables created by the objective variable creating unit 52 . For example, the prediction model creation unit 53 internally joins the monthly explanatory variable (merged version) for each individual shown in FIG. 8 and the monthly objective variable for each individual shown in FIG. Create training data as shown. More specifically, the predictive model creation unit 53, for example, combines the explanatory variables for the same user from January to June 2017 and the objective variable for the prediction implementation month of July 2017. , to create learning data for which the prediction execution month is July 2017. Also, for example, by combining columns of explanatory variables for the same user from February to July 2017 and target variables for prediction execution month August 2017, learning data for prediction execution month August 2017 to create In the learning data shown in FIG. 10, the user ID and prediction execution month information are used in the above-described column combination, and are not used in generating the prediction model.

Next, in step S16 of FIG. 4, the prediction model creating unit 53 creates a prediction model using the learning data, and stores the created prediction model in the prediction model DB 60. FIG. Specifically, the predictive model creating unit 53 uses the learning data as an input and uses machine learning (binary classification method) to create a predictive model that predicts the objective variable from the explanatory variable group. As a machine learning method, for example, logistic regression, decision tree, random forest, SVM (Support Vector Machine), etc. can be used. The prediction model creating unit 53 stores the created prediction model in the prediction model DB 60 . Although the format of the prediction model created differs depending on the machine learning method used, in most machine learning, the degree of importance that indicates how much each explanatory variable contributes to prediction from the internal parameters of the model can be calculated. The predictive model creating unit 53 stores the calculated importance of explanatory variables in an importance table 62 as shown in FIG. 11 . The prediction model created by the prediction model creation unit 53 is used by the prediction model application unit 54 to calculate the prediction result, and the importance is set to visualize the prediction basis when the output unit 55 outputs the prediction result. used.

(Regarding the processing of the predictive model application unit 54)
Next, the processing of the prediction model application unit 54 will be described in detail along the flowchart of FIG. 12 . The processing in FIG. 12 is assumed to be executed at the timing when new data is obtained (for example, once a month) or at any period or timing specified by the user.

In step S110, the prediction model application unit 54 creates explanatory variables using the newly obtained data. FIG. 13 shows an example of explanatory variables created from newly obtained data. The explanatory variables in FIG. 13 are for calculating the risk score in October 2018. By aggregating the data of each user from April to September 2018, monthly explanatory variables for each individual (Merge version) Created in the same manner as (FIG. 8).

Next, in step S112, the prediction model application unit 54 applies the prediction model to the created explanatory variables to calculate the risk score. Specifically, the predictive model applying unit 54 reads the predictive model created by the predictive model creating unit 53 in step S16 of FIG. 4 from the predictive model DB 60 . Then, the predictive model application unit 54 applies the read predictive model to the newly created explanatory variables, thereby calculating a risk score indicating the possibility of each user falling into a problematic situation. FIG. 14( a ) shows the risk score of each user, which is the prediction result of the prediction model application unit 54 . The prediction model application unit 54 transmits the prediction result of FIG. 14( a ) to the output unit 55 . Note that the prediction model application unit 54 also transmits the explanatory variables in FIG. 13 to the output unit 55 together with the prediction result.

(Regarding the process (part 1) of the output unit 55)
Next, the processing (part 1) of the output unit 55 will be described in detail along the flowchart of FIG. 15 . The process of FIG. 15 is executed by the output unit 55 when a request for display of a list screen relating to gambling addiction is input by an employee of a gambling addiction countermeasure organization or the like.

In the process of FIG. 15, first, in step S120, the output unit 55 creates data for screen creation. Specifically, the output unit 55 executes processing according to the flowchart of FIG. 16 . In step S140 of FIG. 16, the output unit 55 calculates the z-score for each explanatory variable of each user acquired from the prediction model application unit 54, and creates a z-score table. A z-score means a value obtained by normalizing one explanatory variable so that the average of all users is 0 and the variance is 1. The z-score is an index representing how much each user's explanatory variable differs from the standard. FIG. 17(a) shows a z-score table created based on the explanatory variables in FIG. It can be said that the z-score of each explanatory variable is information on gambling behavior corresponding to each of a plurality of items (explanatory variables) detected for each user.

Next, in step S142, the output unit 55 calculates z-score*importance (=contribution) to create a contribution table. Specifically, the output unit 55 multiplies the z-score of each explanatory variable for each user in FIG. 17A by the importance of each explanatory variable stored in the importance table 62 in FIG. The contribution of explanatory variables is calculated, and a contribution table as shown in FIG. 17(b) is created. In the contribution degree table, the more deviated from the standard explanatory variable and the higher the degree of importance of the explanatory variable, the higher the contribution degree. That is, the degree of contribution is the degree to which each of a plurality of items (explanatory variables) for determining the degree of dependence on gambling, etc., contributes to the magnitude of the value (risk score) related to the degree of dependence on gambling, etc. of the user. I can say there is. After step S142, the process proceeds to step S122 in FIG.

15, the output unit 55 creates a list screen using the prediction result of the prediction model application unit 54 (FIG. 14(a)) and the contribution degree table (FIG. 17(b)), The list screen is displayed on the display unit 93 or the created list screen is output to the terminal 12 and displayed on the display unit of the terminal 12 . The list screen is a screen as shown in FIG. The output unit 55 displays information on each user's gambling addiction in descending order of risk score on the list screen of FIG. 18 . Here, information of "RANK", "ID", "score", and "simplified analysis result" is displayed on the list screen. "RANK" means the order of risk score, "ID" means user ID, and "score" means risk score value. In the "simplified analysis result" column, a predetermined number (for example, four) of explanatory variables and their contributions are displayed in descending order of contribution for each user.

In this embodiment, by displaying a list screen as shown in FIG. 18, it is possible to display information (risk scores and reasons for prediction) regarding the latest prediction results for multiple users.

Next, in step S124, the output unit 55 waits until the user is selected on the list screen of FIG. FIG. 19 shows an example in which a user with RANK=5 and ID=04946392 is selected. When the user is selected as shown in FIG. 19, the process proceeds to step S126.

After shifting to step S126, the output unit 55 displays the information of the selected user on the list screen. In this case, as shown in the upper part of the list screen in FIG. 19, the output unit 55 outputs predetermined information (risk score, RANK, age, total stay time, , gambling career, gaming information) are extracted from FIG. 13 and the like and displayed. In addition, in the list screen of FIG. 19, the color indicating the user's risk score (=0.74) is changed according to the numerical value of the risk score (see the hatched portion in FIG. 19). This makes it easy to see whether the risk score is high or low.

Returning to FIG. 15, in the next step S128, the output unit 55 waits until the details button corresponding to the selected user is pressed. That is, in the example of FIG. 19, the output unit 55 waits until the button indicated by symbol A is pressed. When the detail button is pressed, the output unit 55 proceeds to step S130.

After shifting to step S130, the output unit 55 executes a process of specifying a countermeasure against gambling addiction of the selected user. In this step S130, the process according to the flowchart of FIG. 20 is executed.

In the process of FIG. 20, first, in step S150, the output unit 55 determines whether or not the risk score is greater than a predetermined threshold. If the determination in step S150 is negative, that is, if the risk score of the selected user is smaller than the threshold, there is no need to present countermeasures, so the process proceeds to step S164, and the output unit 55 Decides not to propose countermeasures. After that, all the processes in FIG. 20 are completed, and the process proceeds to step S132 in FIG.

On the other hand, if the determination in step S150 is affirmative, the process proceeds to step S152, and the output unit 55 selects an explanatory variable that satisfies a predetermined condition from among the selected explanatory variables of the user. In this case, for example, the output unit 55 selects an explanatory variable whose z score x importance (=contribution) is equal to or greater than a threshold value, or selects a predetermined number of explanatory variables that are ranked high in z score x importance (=contribution). to select.

Next, in step S154, the output unit 55 determines whether or not the selected explanatory variables include explanatory variables that can be controlled by human will. Specifically, the output unit 55 refers to the explanatory variable type table shown in FIG. or not. Here, in the table of FIG. 14(b), the number of visits, the bet amount, and the like are explanatory variables that can be controlled by human intention, and the gender is an explanatory variable that cannot be controlled by human intention. is described.

If the determination in step S154 is negative, that is, if the selected explanatory variables do not include any explanatory variables that can be controlled by human will, the process proceeds to step S164. When the process proceeds to step S164, the output unit 55 determines not to present countermeasures, and after that, completes all the processes in FIG. 20, and proceeds to step S132 in FIG.

On the other hand, if the determination in step S154 is affirmative, the process proceeds to step S156, and the output unit 55 selects one explanatory variable that can be controlled by human will. Next, in step S158, the output unit 55 sets the reduction target value based on the average value of people for whom no event has occurred. Next, in step S160, the output unit 55 creates the text of the proposed countermeasure using the preset rule and the preset reduction target value. For example, the output unit 55 outputs text such as "If you reduce the number of days you play by 30% to three times a week, you can reduce the risk of addiction by 40%" If you reduce it to 〇〇 yen, you can reduce the risk of addiction by 20%."

Next, in step S162, the output unit 55 determines whether or not all explanatory variables (explanatory variables controllable by human will) have been selected. If the determination in step S162 is negative, the process returns to step S156, and the processing and determination in steps S156 to S162 are repeatedly executed until the determination in step S162 is affirmed. Then, when the determination in step S162 is affirmative, the output unit 55 completes all the processing in FIG. 20 and proceeds to step S132 in FIG.

After shifting to step S132 in FIG. 15, the output unit 55 displays a detailed screen. Specifically, the output unit 55 displays a detailed screen as shown in FIG.

For example, as shown in FIG. 21, the output unit 55 displays predetermined information (risk score, RANK , age, total time spent, gambling career, gaming information).

In addition, the output unit 55 displays “score transition” in range C of the detailed screen. This score transition column displays the transition of the selected user's past risk score in chronological order.

In addition, the output unit 55 displays information (score basis) regarding the explanatory variables used as the basis for calculating the risk score in the range D of the detailed screen. In this case, as shown in FIG. 21, the output unit 55 displays explanatory variables in descending order of contribution (=z score×importance). In addition, the output unit 55 displays the actual value itself and a graph showing the magnitude of the value in the actual value column. In addition, the output unit 55 represents the maximum actual value by the position of the right end of the graph, represents the average of able-bodied people by white circles (◯), and represents the average of gambling addicts by black circles (●). In addition, the output unit 55 displays the value of importance itself and a graph showing the magnitude of the value in the column of importance, and displays the value of z score x importance (=contribution) in the column of z score x importance. Display the value itself.

In addition, the output unit 55 displays a "countermeasure" button in the area D of the detailed screen in association with the explanatory variable for which the text of the countermeasure was created in step S130 (the process of FIG. 20) described above. When this countermeasure button is pressed, as shown in FIG. 22, the output unit 55 pops up the text of the countermeasure corresponding to the pressed button.

In this embodiment, detailed screens such as those shown in FIGS. 21 and 22 are displayed on the display unit 93 or the display unit of the terminal 12, so that detailed information on the gambling addiction of the selected user can be confirmed. becomes. In addition, since the score transition (range C) is displayed on the detailed screen, the user can easily check whether or not the risk of gambling addiction is increasing. In addition, it is possible to easily confirm the position of the selected user in the overall distribution of actual values, the importance of each explanatory variable, and the degree of influence each explanatory variable has on the risk score.

In addition, by pressing a countermeasure button, a rule-based countermeasure is displayed, so it is possible to confirm what kind of countermeasure should be taken to reduce the risk of gambling addiction.

Therefore, a user such as an employee of a gambling addiction countermeasure organization can identify a user at risk of gambling addiction by referring to the screens of FIGS. It is possible to notify external institutions of the reasons for high risk scores and appropriate countermeasures.

In the processing of FIG. 15, a case has been described in which countermeasures are specified before displaying the detailed screen (S130) and countermeasure buttons are displayed on the detailed screen (S132). However, it is not limited to this. For example, at the timing when a certain explanatory variable is selected from the detailed screen, a countermeasure related to the selected explanatory variable is identified, and if it is identified, the countermeasure may be displayed.

In addition, although FIG. 21 and FIG. 22 explained the case where the explanatory variables are displayed in descending order of contribution, the explanatory variables may be displayed in descending order of contribution. Further, the display order of the explanatory variables may be switched between the descending order of contribution and the descending order of contribution. In addition, in FIGS. 21 and 22, all explanatory variables may be displayed, or a part of explanatory variables (a predetermined number of them in descending order of contribution) may be displayed.

(Regarding the process (part 2) of the output unit 55)
Next, the processing (part 2) of the output unit 55 will be described along the flowchart of FIG. 23 and with appropriate reference to other drawings. The processing of FIG. 23 is executed by the output unit 55 when a request to display the score transition screen for multiple users (the screen of FIG. 25) is input by an employee of a gambling addiction countermeasure organization.

In the process of FIG. 23, first, in step S200, the output unit 55 acquires the risk score of each user for each predetermined period. Here, as an example, the output unit 55 acquires the risk score of each user for each week. That is, the prediction model application unit 54 is assumed to perform prediction processing (FIG. 12) every week.

Next, in step S202, when the output unit 55 displays the obtained risk scores for each user in chronological order as shown in FIG. 25, a method of emphasizing the background color within the display frame is specified. In this case, the output unit 55 specifies the background color enhancement method based on, for example, an enhancement method table corresponding to the risk score as shown in FIG. 24(a). Here, in the table of FIG. 24(a), an emphasis method (background color) is defined for each risk score range. Therefore, the output unit 55 identifies the background color of each risk score obtained in step S200 by referring to the table of FIG. 24(a).

Next, in step S204 of FIG. 23, the output unit 55 specifies the event occurrence date of each user. In this case, the output unit 55 identifies the event occurrence date of each user from the event information of FIG. 9(c). When creating the target variable for machine learning in step S12 described above, if it is included in the period from the event occurrence date to a certain period before that event, it is treated as a "positive example", so that the event occurrence can be predicted in advance. Realized. Therefore, for the purpose of confirming the certainty of the advance prediction, in the following description, instead of the event occurrence date, the date a certain period before the event occurrence date may be used.

Next, in step S206, the output unit 55 identifies an emphasis method based on the degree of risk determined from the event occurrence date of each user and the change trend (or stability) of the risk score. In this case, the output unit 55 specifies the emphasizing method based on, for example, an emphasizing method table corresponding to the degree of risk as shown in FIG. 24(b). Here, as shown in FIG. 24B, the degree of risk is determined for each combination of the change trend (inclination) of the risk score and the event occurrence situation. For example, the output unit 55 sets the moving window period to four weeks, and identifies the degree of risk for each week included in the moving window period based on the change trend (inclination) of the risk score within the moving window period and the event occurrence status. The emphasizing method corresponding to the calculated degree of risk is identified from the table of FIG. 24(b). As the slope, a slope obtained by regression analysis or the like can be used, or a moving average of slopes in a moving window period can be used. For example, if the slope of the score increase in the moving window period is 2.0 or more and the week included in the moving window period is the week after the event occurs, the output unit 55 encloses the week in a square. , to display the text indicating the risk score in pink and bold. In the present embodiment, the date of event occurrence of each user identified in step S204 by the output unit 55 is information associated with each user, and the risk determined in step S206 is the risk associated with gambling addiction of each user. degree.

Next, in step S208, the output unit 55 generates a screen displaying the risk score of each user for each predetermined period acquired in step S200 based on the emphasis method specified in steps S202 and S206, and displays the display unit 93 and It is displayed on the display unit of the terminal 12 .

FIG. 25 shows an example of the screen displayed in step S208. On the screen of FIG. 25, the weekly risk score (value) of each user is displayed in chronological order. Moreover, on the screen of FIG. 25, the background color is color-coded according to the risk score. In addition, on the screen of FIG. 25, locations where the risk score changes rapidly are enclosed by squares. In addition, in FIG. 25, when two or more adjacent risk scores are enclosed in a square, the two or more adjacent risk scores are enclosed in one square. Furthermore, on the screen of FIG. 25, the risk score (value) for the period after the occurrence of the event is indicated in bold (in FIG. 25, the characters are displayed large for convenience of illustration). Although not shown on the screen of FIG. 25, the character colors are also color-coded based on the table of FIG. 24(b).

According to the screen of FIG. 25, focusing on the row of the arrow E, it can be seen that there are locations where the risk score is displayed in bold letters even though there are no locations surrounded by squares. In this case, there are users for whom events have occurred even though their risk scores are not on an upward trend. In such a case, the staff of the gambling addiction countermeasures agency, etc. can judge that prediction omissions have occurred and that the prediction model needs to be improved. Also, when focusing on the row of the arrow F, it can be seen that the risk score of the portion surrounded by the square is displayed in bold. In this case, since the risk score has increased and an event has occurred, the staff of the agency for gambling addiction countermeasures can judge that the prediction is appropriate. Also, when focusing on the row of the arrow G, it can be seen that there are portions surrounded by squares where the risk score is not displayed in bold. In this way, if there are users whose risk scores have risen but no events have occurred, officials of organizations dealing with gambling addiction should pay special attention to such users in the future. It can be determined that it is the user. In addition, it can be judged that the staff of the gambling addiction countermeasures agency, etc., have failed to respond to the target users, and that it is necessary to improve the business process such as the standard value of event occurrence.

In this manner, in the present embodiment, changes in the risk score of each user and the presence or absence of event occurrence can be displayed on the screen in an easy-to-understand manner. Also, by referring to the screen, it becomes possible to discover the necessity of improving the prediction model.

Here, in this embodiment, display control is performed so that the background color and the character color do not match. This makes it possible to display the risk score in an easy-to-read manner.

Note that, in the present embodiment, the range surrounded by a rectangle can also be a range in which the risk score has continuously increased by a threshold value or more (for example, three times or more) compared to the previous week. Also, the range enclosed by the rectangle can be a range in which the score increases compared to the previous time at a rate equal to or higher than a threshold (for example, 75% or higher).

In this embodiment, the background color is controlled based on the value of the risk score, and the color and thickness of the characters are controlled based on the degree of risk. However, the present invention is not limited to this. That is, the display mode of either the background color or the characters may be changed based on the value of the risk score, and the display mode of either the background color or the characters may be changed based on the degree of risk. When changing the display mode of characters, the size of the characters may be changed.

In this embodiment, as shown in FIG. 26, the output unit 55 changes the background color based on the range of the risk score, and also changes the character of the risk score for the week in which the event occurred and the week after that. It is also possible to simply change the display mode (for example, make it bold). Even in this way, it is possible to easily discover that there is a discrepancy between the prediction and the actual result, and the period in which the discrepancy exists. This makes it easier for staff members of gambling addiction countermeasures organizations to determine the necessity of improving the prediction model. In addition, it is possible for the staff of the gambling addiction countermeasures organization to easily find exceptional users such as high-income earners who play games frequently and at a high cost but who do not fall under the category of addiction.

After displaying the screens shown in FIGS. 25 and 26, the output unit 55 displays a detailed screen of the user as shown in FIG. 21 when a user (user ID) is selected on the screen. may

As described in detail above, according to the present embodiment, the output unit 55 acquires the risk score calculated by the prediction model application unit 54, and the z-score of the explanatory variable and the importance level stored in the importance level table 62. Contribution is calculated based on and. Then, the output unit 55 displays the explanatory variables in descending order of contribution on the detailed screen (FIGS. 21 and 22). This makes it possible to clearly display the explanatory variables that affect the magnitude of the risk score value of each user.

Further, according to the present embodiment, the output unit 55 displays the risk score on the detailed screen (FIGS. 21 and 22), and also displays the time-series transition of the risk score. This makes it possible to display the value of the user's risk score and the transition of the risk score in an easy-to-understand manner.

In addition, in the present embodiment, the output unit 55 creates and displays countermeasures against addiction to gambling or the like based on the risk score and the degree of contribution. This makes it possible to automatically create and display appropriate countermeasures based on risk scores and contributions.

In addition, the output unit 55 of the present embodiment determines whether or not to create countermeasures based on the type of explanatory variable (whether or not the explanatory variable is controllable by human will) (S154). This makes it possible to appropriately select explanatory variables for which countermeasures need to be created.

In addition, in this embodiment, the output unit 55 controls the background color of the area where the risk score is displayed according to the risk score when displaying the screens of FIGS. 25 and 26 . In addition, the output unit 55 controls the color and thickness of the letters of the risk score according to the degree of risk determined based on the risk score and whether or not an event has occurred. This makes it possible to display the relationship between the magnitude of the risk score and the degree of danger in an easy-to-understand manner. In this case, the staff of the gambling addiction countermeasures agency, etc. will judge the necessity of improving the prediction model based on the relationship between the magnitude of the risk score and the magnitude of the risk, and identify users who should be noted. can be

In the above embodiment, the server 80 functions as an ICT infrastructure and performs the above-described processes, but the present invention is not limited to this. Other servers or devices, such as the server 30 managed by a gambling addiction countermeasures organization and the server 10 managed by a public institution, may function as the ICT infrastructure and execute the above-described processing.

In the above embodiment, a device other than the server 80 included in the information processing system 100 may have functions other than the output unit 55 of the server 80 (see FIG. 3). Furthermore, if a device other than the server 80 included in the information processing system 100 has a part of the functions of the output unit 55 in the above-described embodiment (for example, the function of executing the process of step S120 in FIG. 15). good too.

Note that the processing functions described above can be realized by a computer. In that case, a program is provided that describes the processing contents of the functions that the processing device should have. By executing the program on a computer, the above processing functions are realized on the computer. A program describing the processing content can be recorded in a computer-readable storage medium (excluding carrier waves).

When a program is distributed, it is sold in the form of a portable storage medium such as a DVD (Digital Versatile Disc) or a CD-ROM (Compact Disc Read Only Memory) on which the program is recorded. It is also possible to store the program in the storage device of the server computer and transfer the program from the server computer to another computer via the network.

A computer that executes a program stores, for example, a program recorded on a portable storage medium or a program transferred from a server computer in its own storage device. The computer then reads the program from its own storage device and executes processing according to the program. The computer can also read the program directly from the portable storage medium and execute processing according to the program. In addition, the computer can also execute processing in accordance with the received program each time the program is transferred from the server computer.

The embodiments described above are examples of preferred implementations of the present invention. However, the present invention is not limited to this, and various modifications can be made without departing from the spirit of the present invention.

In addition, the following additional remarks will be disclosed with respect to the above description of the embodiment.
(Appendix 1) an acquisition unit that acquires information about actions corresponding to each of a plurality of items detected for a certain person;
Based on the information on the behavior acquired by the acquisition unit and the weight information of each of the plurality of items stored in the storage unit, a value related to the degree of dependence of the person on a certain object and a degree of dependence on the certain object are determined. a calculating unit for calculating a degree of contribution indicating the degree to which each of the plurality of items for performing the calculation contributes to the magnitude of the value;
a display control unit that displays some or all of the items included in the plurality of items in an order according to the degree of contribution calculated by the calculation unit;
An information processing system characterized by having:
(Supplementary note 2) The information processing system according to Supplementary note 1, wherein the display control unit also displays information on the value and/or the transition of the value along a time series.
(Supplementary note 3) The information processing system according to Supplementary note 1 or 2, wherein the display control unit further determines and displays a countermeasure regarding dependence on the certain target based on the value and the contribution degree. .
(Supplementary note 4) The information processing program according to Supplementary note 1 or 2, wherein the display control unit further determines and displays countermeasures for a certain item for which designation is received.
(Supplementary note 5) Supplementary note 1 characterized in that the display control unit determines, among the plurality of items, an item for which countermeasures are to be displayed, according to the information of the type associated with each of the plurality of items. 5. The information processing program according to any one of 1 to 4.
(Appendix 6) Acquiring information about behavior corresponding to each of a plurality of items detected for a certain person,
Based on the obtained information about the action and the weight information of each of the plurality of items stored in the storage unit, the Calculate a contribution degree indicating the degree to which each of the plurality of items contributes to the magnitude of the value,
displaying some or all of the items included in the plurality of items in an order according to the calculated degree of contribution;
An information processing method characterized in that a computer executes processing.
(Appendix 7) Dependence degree of a certain person on a certain target calculated based on information on actions corresponding to each of a plurality of items detected for a certain person and weight information of each of the plurality of items stored in the storage unit and a contribution indicating the degree to which each of the plurality of items for determining the degree of dependence on the certain object contributes to the magnitude of the value;
An information processing program for causing a computer to execute a process of displaying some or all of the items included in the plurality of items in an order according to the acquired contribution value.
(Appendix 8) The process of displaying some or all of the items included in the plurality of items in the order according to the degree of contribution includes information on the value and/or the transition of the value along the time series. The information processing program according to appendix 7, further displaying
(Supplementary Note 9) The process of displaying some or all of the items included in the plurality of items in the order according to the degree of contribution is further based on the value and the degree of contribution. 9. The information processing program according to appendix 7 or 8, which determines and displays countermeasures for the above.
(Supplementary Note 10) The process of displaying some or all of the items included in the plurality of items in the order according to the degree of contribution further includes determining and displaying countermeasures for the specified item. The information processing program according to appendix 7 or 8, characterized by:
(Supplementary Note 11) The process of displaying some or all of the items included in the plurality of items in the order according to the degree of contribution is performed according to the type information associated with each of the plurality of items. 11. The information processing program according to any one of appendices 7 to 10, wherein an item for displaying countermeasures is determined from among the plurality of items.
(Supplementary Note 12) Dependence of a certain person on a certain target calculated based on information on actions corresponding to each of a plurality of items detected for a certain person and weight information of each of the plurality of items stored in a storage unit and a contribution degree indicating the degree to which each of the plurality of items for determining the degree of dependence on the object contributes to the magnitude of the value;
a display control unit that displays some or all of the items included in the plurality of items in an order according to the magnitude of the contribution value acquired by the acquisition unit;
An information processing device comprising:

55 output unit (acquisition unit, calculation unit, display control unit)
80 server (information processing device)
100 information processing system

Claims (8)

an acquisition unit that acquires an index indicating how much information about behavior for each of a plurality of items about a person differs from a standard ;
obtaining the degree of contribution indicating the degree of contribution of each of the plurality of items to the magnitude of the value indicating the degree of dependence of the person on the target obtained from the information on the behavior of each of the plurality of items; a calculation unit that calculates by multiplying the index of each of the plurality of items acquired by the unit by weight information of each of the plurality of items stored in a storage unit;
a display control unit that displays some or all of the items included in the plurality of items in an order according to the degree of contribution calculated by the calculation unit;
An information processing system characterized by having: Obtaining an index indicating how much information about behavior for each of a plurality of items about a person differs from a standard ,
Contribution degrees indicating the extent to which each of the plurality of items contributes to the magnitude of the value indicating the degree of dependence of the person on a certain object obtained from the information on the behavior of each of the plurality of items is acquired. calculated by multiplying the index of each of the plurality of items by weight information of each of the plurality of items stored in a storage unit;
displaying some or all of the items included in the plurality of items in an order according to the calculated degree of contribution;
An information processing method characterized in that a computer executes processing. The plurality of items calculated by multiplying the weight information of each of the plurality of items stored in the storage unit by an index indicating how much the information on behavior for each of the plurality of items related to a certain person differs from the standard obtaining a contribution degree indicating the degree of contribution of each of the plurality of items to the magnitude of the value indicating the degree of dependence of the certain person on a certain object obtained from the information on the behavior of each of the items ;
An information processing program for causing a computer to execute a process of displaying some or all of the items included in the plurality of items in an order according to the acquired contribution value. The process of displaying some or all of the items included in the plurality of items in the order according to the degree of contribution may also display information on the value and/or the transition of the value along a time series. The information processing program according to claim 3, characterized by: The process of displaying some or all of the items included in the plurality of items in the order according to the degree of contribution is further based on the value, the degree of contribution , and a preset rule , 5. The information processing program according to claim 3, wherein a text of countermeasures regarding dependence on said certain target is created and displayed. 6. The information processing program according to claim 5 , wherein , when a specification of a certain item is accepted, a text of countermeasures for the specified item is created and displayed. The process of displaying a part or all of the items included in the plurality of items in the order according to the degree of contribution includes: 7. The information processing program according to any one of claims 3 to 6, wherein an item for displaying countermeasures is determined from among the items. The plurality of items calculated by multiplying the weight information of each of the plurality of items stored in the storage unit by an index indicating how much the information on behavior for each of the plurality of items related to a certain person differs from the standard an acquiring unit that acquires a degree of contribution indicating the degree of contribution of each of the plurality of items to the magnitude of the value indicating the degree of dependence of the person on a certain object obtained from the information on the behavior of each of the items of ,
a display control unit that displays some or all of the items included in the plurality of items in an order according to the magnitude of the contribution value acquired by the acquisition unit;
An information processing device comprising:

Images