JP7075173B2 - Programs, devices and methods for determining behavioral changes in user groups - Google Patents

Description

The present invention relates to a technique for determining a behavioral change of a group of users. For example, it is suitable for determining the economic situation of a company from changes in the behavior of the employees of the company.

Business owners are trying to maximize profits, that is, to increase stock dividends and stock price values, by utilizing the people, goods, and money of management resources for the shareholders who are the owners. On the other hand, risk management with external business partners is also important, and we try to collect the management status from the financial statements of business partners.
For companies, good rumors attract new business partners while strengthening relationships with existing business partners. On the other hand, bad rumors spread in a blink of an eye and keep business partners away. This is part of risk management activities as a company and is to avoid losses due to the bankruptcy of business partners.
In a commercial transaction between an individual and a store, the price of the goods is paid in cash at the same time as the goods are delivered. On the other hand, in general business transactions between companies, an invoice is issued after the delivery of goods, and the payment is made, for example, three months later. That is, the company that delivered the goods will "credit" the delivery destination company until the price is paid, and will be in a state of having a bankruptcy risk (credit risk). Credit means believing that the other party will pay for it in the future.

A company uses a "financial statement" issued by the business partner on a regular basis to predict signs of bankruptcy of the business partner. For example, it is possible to machine-learn the financial information of a company that has gone bankrupt in the past and the financial information of a company that has not gone bankrupt, build a discriminant model such as a logistic regression model, and calculate the bankruptcy probability in advance. Such a method is used especially in banks and the like. This contributes to uniform credit management (setting of loan amount and interest rate, setting of collateral, credit protection and recovery rate forecast, general credit risk management) and shortening of examination time. .. These are called "scoring models", and as a well-known method, there is a Merton model using Altman's Z score and stock price.

There are two main cases of bankruptcy or bankruptcy of a company.
(Case 1) Case of gradual decline over a long period of time leading to bankruptcy / bankruptcy (Case 2) Case of sudden deterioration of business performance in a short period of time leading to bankruptcy / bankruptcy

In case 1, there is a high possibility that dangerous signs can be read from the financial indicators of the financial statements. This is because if the bankruptcy gradually progresses, the signs will be gradually reflected in the financial statements. In the case of machine learning, the model builder extracts statistically useful signs. The conventional scoring model is particularly effective in cases such as Case 1, which gradually goes bankrupt.

As another conventional technique, as a credit transaction to an individual, there is a technique of estimating the creditworthiness of each user (see, for example, Patent Document 1). According to this technique, for each user, the demand for credit transactions is estimated, and a plurality of users are classified into a plurality of user groups based on the estimated creditworthiness and demand.
Further, in order to determine the credit information for an individual, the location information of the terminal device used by the user is compared with the location information related to the user information that the user has registered in the predetermined financial service in the past, and the location information thereof is compared. There is also a technique for determining a change in user information based on a change in position information (see, for example, Patent Document 2).

Japanese Patent No. 6472915 Japanese Unexamined Patent Publication No. 2019-049993

"Is it illegal to reduce salary? Three points to note when lowering employee salary", [online], [Search on August 12, 1st year of Reiwa], Internet <URL: https://kigyosapo.com/sdown> "Advance payment business for unpaid wages", [online], [Search on August 12, 1991], Internet <URL: https://www.johas.go.jp/tabid/417/Default.aspx> "What is a two-sample t-test?", [Online], [Search on August 12, 1991], Internet <URL: https://bellcurve.jp/statistics/course/9427.html>

In the case of the quarterly financial statements of listed companies, the financial statements that publicize the economic situation of the companies are issued only four times a year. In addition, the financial statements will be issued about two months after the record date. Therefore, as in case 2 described above, if the business performance suddenly deteriorates in a short period of time, the financial statements will be published with a delay, which is a fatal result.

As another analysis method in corporate evaluation, bankruptcy can be predicted, for example, from deposit / withdrawal information of a bank account. However, companies can also manipulate deposit and withdrawal information to hide the deterioration of economic conditions. Simply, bank A's account a is for submission and is repaired so that deposits and withdrawals can be considered normal. On the other hand, the account b of the bank B is a flow of funds that reflects the actual situation of the company. In this case, it is necessary to know the deposit / withdrawal information of all the bank accounts held by the company.

If the economic situation of a company is deteriorating, in either case 1 or 2, the management of the company first improves the situation while hiding the economic situation from other business partners. try to. That is, before applying for a loan to a financial institution, seek another method.
Companies consider it the last resort to request financial support from their business partners, even if the economic situation is critical. This is because the business partner is a negotiating partner and may have weaknesses. It would be nice if the bank as a financial institution could provide immediate support, but the bank's examination always tries to bring out the negative side. When a business owner requests financial assistance from a bank, the bank will first raise interest rates on existing loans and require the submission of any collateral, including real estate holdings. Therefore, even if the economic situation is critical, the manager of the company tries to improve the economic situation only within his / her own company.

Business owners first try to take measures such as cost reduction. Next, as a measure that can impose a burden unilaterally, we try to change the salary payment conditions to employees under the constraint condition that information is not leaked to other business partners.
Changes in salary payment conditions include a reduction in salary and a delay in payment due. It is said that it is legal to reduce salary up to about 10% when the salary is reduced for rational reasons such as deterioration of economic conditions.
According to Article 91 of the Labor Standards Law, if the rules of employment stipulate that a worker's salary will be reduced by sanctions, the amount of each reduction will exceed half of the average wage per day, and the total amount will be paid by one wage. It is stipulated that the wage must not exceed 1/10 of the wage in the period. As for the administrative procedure, the employee is informed of the salary reduction plan, then the work rules are changed, and the employee submits a consent form. This legally changes salary terms and conditions. What is important here is that a company can legally take such measures without publicly disclosing the deterioration of changes in economic conditions (see, for example, Non-Patent Document 1).

In addition, according to the Workers' Health and Safety Organization (an affiliated organization of the Ministry of Health, Labor and Welfare), the number of reimbursement of unpaid wages of companies in 2016 is 8.4 billion yen for 21,941 people (for example, Non-Patent Document 2). reference). That is, it can be confirmed that the salary payment to the employee is delayed before the bankruptcy.

Employees whose salary payment terms have changed need to live in a way that accommodates reduced salaries and late payments. Specifically, they may refrain from costly behavior until their salary is repaid, reduce their living expenses, and cut down on their own savings. Employees will first stop visiting costly places, such as outing leisure, although costs such as clothing, food and shelter cannot be easily reduced. On the other hand, it may be easier to visit if it is a low-cost destination, even if you are outing.

On the other hand, the inventors of the present application may be able to determine the change in the economic situation of the company if the behavior change of the employee (user) group of the company as a whole can be determined. Thought. Employees, in particular, believe that in the event of a salary reduction or payment delay from a company, they will first try to avoid wasteful travel and visits to expensive locations. rice field.

Therefore, an object of the present invention is to provide a program, an apparatus, and a method for determining a behavioral change of a group of users.

According to the present invention, it is a program for operating a computer mounted on a device for determining a behavioral change of a group of users.
A user position database that accumulates time-series positions on mobile terminals owned by each user in the user group, and
An estimated payment amount database that stores the estimated payment amount in association with each region,
For each user, for each predetermined period, the user location database is used to identify one or more areas of stay, and the estimated payment amount database is used to add up the estimated payment amounts of all areas of stay within the specified period. A means of calculating the total expected payment amount to calculate the amount, and
A user behavior change determining means for determining whether or not the total expected payment amount has changed to a predetermined amount or more according to the elapse of a predetermined period for each user.
When the ratio of users determined to be true by the user behavior change determination means exceeds a predetermined ratio, the computer functions as a user group behavior change determination means that activates that a behavior change has occurred in the user group. It is a feature.

According to other embodiments in the program of the invention
The user group is the employee group of the company,
It is also preferable to further function the computer so that the occurrence of a change in the behavior of the user group means the occurrence of a change in the economic situation of the company.

According to other embodiments in the program of the invention
It is also preferable that the means for calculating the total expected payment amount has the computer function so as to extract only the area where the user has stayed for a predetermined time or longer as the area of stay.

According to other embodiments in the program of the invention
The user behavior change determination means may also make the computer function so as to determine whether or not the total expected payment amount in the previous predetermined period has changed to a predetermined amount or more than the total expected payment amount in the later predetermined period. preferable.

According to other embodiments in the program of the invention
The expected payment amount database associates the average amount and standard deviation as the expected payment amount for each region.
The total expected payment amount calculation means calculates, for each user, the total expected payment amount obtained by adding the average amount of the expected payment amount and the total expected standard deviation which is the square root of the added value of the squared (variance) of the standard deviation.
It is also preferable that the user behavior change determining means make the computer function so as to determine whether or not the total expected payment amount has changed to a predetermined amount or more, or the total expected standard deviation has changed to a predetermined standard deviation or more.

According to other embodiments in the program of the invention
The user behavior change determination means is based on a two-sample t-test.
The F-test of the ratio of the population variance determines whether the null hypothesis of homoscedasticity is rejected or rejected.
If the null hypothesis of homoscedasticity is not rejected, the presence or absence of a significant difference is determined by the t-test assuming that the variances are similar.
If the null hypothesis of homoscedasticity is rejected, it is also preferable to have the computer function to determine the presence or absence of a significant difference by Welch's t-test, which assumes that the variances are different.

According to other embodiments in the program of the invention
With respect to the user position database, the time-series positions are such that the terminal positioning position is positioned by the mobile terminal possessed by the user and / or the base station positioning position based on the stay range of the mobile terminal connected to the base station. It is also preferable to make the computer work.

According to the present invention, it is a program for operating a computer mounted on a device for determining a change in a company's economic situation.
For each user who is an employee of the company, the total expected payment amount that adds up each expected payment amount and the total expected standard deviation that is the square root of the variance of the expected payment amount is calculated for each predetermined period. Calculation means and
As a user behavior change determining means for determining whether or not the total expected payment amount has changed to a predetermined amount or more or the total expected standard deviation has changed to a predetermined standard deviation or more according to the elapse of a predetermined period for each user. ,
When the ratio of users determined to be true by the user behavior change determination means exceeds a predetermined ratio, the computer functions as a user group behavior change determination means that activates that a behavior change has occurred in the user group. It is a feature.

According to the present invention, it is a device for determining a behavioral change of a user group.
A user position database that accumulates time-series positions on mobile terminals owned by each user in the user group, and
An estimated payment amount database that stores the estimated payment amount in association with each region,
For each user, for each predetermined period, the user location database is used to identify one or more areas of stay, and the estimated payment amount database is used to add up the estimated payment amounts of all areas of stay within the specified period. A means of calculating the total expected payment amount to calculate the amount, and
A user behavior change determining means for determining whether or not the total expected payment amount has changed to a predetermined amount or more according to the elapse of a predetermined period for each user.
It is characterized by having a user group behavior change determining means that activates that a behavior change has occurred in the user group when the ratio of users determined to be true by the user behavior change determining means becomes a predetermined ratio or more. do.

According to the present invention, it is a device for determining a change in the economic situation of a company.
For each user who is an employee of the company, the total expected payment amount that adds up each expected payment amount and the total expected standard deviation that is the square root of the variance of the expected payment amount is calculated for each predetermined period. Calculation means and
As a user behavior change determining means for determining whether or not the total expected payment amount has changed to a predetermined amount or more or the total expected standard deviation has changed to a predetermined standard deviation or more according to the elapse of a predetermined period for each user. ,
It is characterized by having a user group behavior change determining means that activates that a behavior change has occurred in the user group when the ratio of users determined to be true by the user behavior change determining means becomes a predetermined ratio or more. do.

According to the present invention, it is a method for determining a change in behavior of a user group of a device.
The device is
A user position database that accumulates time-series positions on mobile terminals owned by each user in the user group, and
An estimated payment amount database that stores the estimated payment amount in association with each region,
Have,
The device is
For each user, for each predetermined period, the user location database is used to identify one or more areas of stay, and the estimated payment amount database is used to add up the estimated payment amounts of all areas of stay within the specified period. The first step to calculate the amount and
For each user, the second step of determining whether or not the total expected payment amount has changed to the predetermined amount or more according to the lapse of a predetermined period, and
It is characterized in that when the ratio of users determined to be true by the second step becomes a predetermined ratio or more, the third step of invoking that the behavior change of the user group has occurred is executed.

According to the present invention, it is a method for determining a change in the corporate economy of an apparatus.
The first step to calculate the total expected payment amount, which is the sum of the expected payment amounts, and the total expected standard deviation, which is the square root of the variance of the expected payment amount, for each user who is an employee of the company. When,
For each user, the second step of determining whether the total expected payment amount has changed to the predetermined amount or more or the total expected standard deviation has changed to the predetermined standard deviation or more according to the lapse of a predetermined period.
It is characterized by executing the third step of invoking that a change in the economic situation of a company has occurred when the ratio of users determined to be true by the second step exceeds a predetermined ratio. ..

According to the program, apparatus and method of the present invention, it is possible to determine the behavioral change of the user group.

It is explanatory drawing which shows the behavior change of the user in this invention. It is a functional block diagram of the estimation apparatus in this invention. It is explanatory drawing which associated with the expected payment amount for each region. It is explanatory drawing which associated with the expected payment amount for each area which is adjacent in a mesh shape. It is 1st explanatory diagram which calculates the total expected payment amount of a user. It is a 2nd explanatory diagram which calculates the total expected payment amount of a user. It is explanatory drawing which shows the process of a user behavior change determination part.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is an explanatory diagram showing a change in user behavior in the present invention.

According to FIG. 1 (a), a plurality of areas M1 to M5 visited by the user from his / her home in the address area are shown. In general, every time a user moves to any area, they will make a visit that incurs some payment charges.
According to FIG. 1A, the average payment amount is associated with each region as follows.
Region M1: Average amount 1,000 yen Region M2: Average amount 200 yen Region M3: Average amount 5,000 yen Region M4: Average amount 20,000 yen Region M5: Average amount 50,000 yen

If the user is an employee of the company, he / she needs to live according to his / her monthly salary. Therefore, for the user, it is inevitably assumed that there is some correlation between the salary amount, the place where the payment fee is incurred, and the number of visits. As shown in FIG. 1 (a), for example, a user with a relatively high salary also visits areas M4 and M5 where the payment fee is relatively high.

On the other hand, if the salary from the company is reduced or the payment is delayed for the user, the expenditure must be reduced as a matter of course. Users will initially try to avoid wasteful travel and visits to expensive locations other than food, clothing and shelter. According to FIG. 1 (b), the user refrains from visiting the areas M4 and M5 where the payment fee is relatively high. In this way, it can be assumed that the reduction of salary causes an indirect change in the behavior of the user.

On the other hand, the inventors of the present application may indirectly increase the rate of some change in the behavior of all employees when the deterioration of the economic situation of the company leads to the reduction of the salary of the employees. I wondered if there was one. When the rate of this change in user behavior exceeds a predetermined rate, it can be assumed that the economic situation of the company deteriorates. As a result, I think that the deterioration of the economic situation of the company can be recognized in advance by business partners, financial institutions, investors, and credit research companies (Tokyo Shoko Research and Teikoku Data Bank). rice field.
If the organization of a company is likened to a pyramid, the president is at the top of the pyramid, and the employees support the foundation. In fact, the business situation that only the top corporate manager can know will change as the behavior of the employees at the bottom (referred to as the "bottom-up approach").

FIG. 2 is a functional configuration diagram of the estimation device in the present invention.

The estimation device 1 determines the behavioral change of the user group. Specifically, the estimation device 1 determines that there is a possibility that some change has occurred in the economic situation of the company when the behavioral change of the employee (user) group of the company occurs.

The estimation device 1 includes a user position database 11, an estimated payment amount database 12, a total expected payment amount calculation unit 13, a user behavior change determination unit 14, and a user group behavior change determination unit 15. These functional components are realized as a program for operating the computer mounted on the device. In addition, the processing flow of these functional components can be understood as a method for determining a behavior change of a user group.

[User location database 11]
The user position database 11 accumulates time-series positions for each user who has a mobile terminal capable of acquiring a position. Here, the target user group is a group of employees of a company.
According to FIG. 2, the user location database 11 is operated and managed by a specific telecommunications carrier, and the time and location information are stored in association with each subscriber ID (user ID).
Here, as an important point, the user position database 11 is a captured user by the communication business equipment of a specific communication carrier, and may not be all the actual users. Of course, the larger the number of populations, the better, but the location information may be collected only for the users who hold the mobile terminal 2 contracted with a specific telecommunications carrier.

In the user position database 11, it is preferable, but not limited to, the employees of the company are specified in advance as the user group. For example, if it is a user who visits the company every day, it may be a non-regular employee, a part-time job, a subcontractor employee, or it may be included as a user group. In the present invention, it is not a problem that the amount of payment generated at the visited place due to the behavior of the user is not a problem, and the time when the behavior change occurs in the entire group of users is determined.

The positions accumulated in the user position database 11 are, for example, as follows.
(1) Terminal positioning position positioned by the mobile terminal 2 possessed by the user This is latitude / longitude information positioned by the mobile terminal 2 itself by GPS (Global Positioning System).
(2) Base station positioning position of a mobile terminal connected to a communication carrier's base station or access point The position of the mobile terminal 2 is estimated from the position information of the base station or access point under the mobile terminal 2. You may. However, this position information has a coarse spatial particle size.
These position information may be represented by latitude / longitude or map coordinates, or may be converted into an address name or a map mesh number.

[Expected payment amount database 12]
The expected payment amount database 12 stores the expected payment amount in association with each other for each region.
When the area Mi is a store, the expected payment amount is collected on a user-by-user basis from POS (Point Of Sale) information and cashless payment information.
The expected payment amount per unit time may be used. In that case, the actual expected payment amount can be predicted by the staying time of the user.

It is also preferable to associate "average amount" and "standard deviation" as expected payment amounts for each region.
As for "region", the more finely divided the flat mesh on the map is, the more the building or facility can be specified. Further, the "region" may be divided into multidimensional spaces such as the number of floors of the building or facility. For example, when dividing in three dimensions, the space can also be called a "cube".
The "average amount" represents the average value of payment amounts paid by many users in the past in the area.
The "standard deviation" refers to the variation in payment amounts paid by many users in the past in the region.

FIG. 3 is an explanatory diagram in which the expected payment amount is associated with each region.

According to FIG. 3, a normal distribution graph of the expected payment amount is associated with each region Mi. The normal distribution can statistically represent past payments as a probability distribution for a continuous variable that represents the distribution of payments that accumulate near the average amount.

FIG. 4 is an explanatory diagram in which expected payment amounts are associated with each region adjacent to each other in a mesh shape.

According to FIG. 4, regions M1, M2, and M3 are adjacent to each other. It is also far from regions M4 and M5 with respect to regions M1, M2 and M3. In this way, the normal distribution of the expected payment amount may be expressed on the map. In general, some regions have relatively high payments, while others have relatively low payments.

It is also preferable to create a payment amount map in which the gradation of color tones is displayed darker (or lighter) according to the payment amount for each predetermined time zone and for each region. As a result, the payment amount can be recognized only by looking at the color of the payment amount map.

[Total expected payment amount calculation unit 13]
The total expected payment amount calculation unit 13 specifies one or more stay areas for each user by using the user position database 11 for each predetermined period. Then, for each user, the total expected payment amount obtained by adding the average amount of the expected payment amounts and the total expected standard deviation which is the square root of the added value of the square (variance) of the standard deviations are calculated.

FIG. 5 is a first explanatory diagram for calculating the total expected payment amount of the user.

According to FIG. 5, the “stay area” detected in a predetermined period (for example, one week) is derived for each user. At this time, the total expected payment amount calculation unit 13 extracts only the area where the user has stayed for a predetermined time (for example, 1 hour) or more as a “stay area”. If the user stays in the area (store) for, for example, for one hour or more, it is assumed that the user has paid the expected payment amount corresponding to the area.
According to FIG. 5, the user ID1 visits the area M2 twice and the area M3 only once during a predetermined period (a week from August 18 to 24).

According to FIGS. 1 to 5 described above, the total expected payment amount and the total expected standard deviation are calculated for each user who is an employee of the company for each predetermined period.
If the total expected payment amount and the total expected standard deviation can be calculated for each predetermined period from the deposit / withdrawal status of the user's bank account regardless of the user's position, the user position database 11 and the expected payment amount database 12 is not an essential component.
In that case, for the user, each payment amount in each stay area can be considered as each withdrawal amount in the user's bank account.

FIG. 6 is a second explanatory diagram for calculating the total expected payment amount of the user.

According to FIG. 6, the total estimated payment amount calculation unit 13 uses the estimated payment amount database 12 for each user to add up the expected payment amounts of all the stay areas within a predetermined period, and the total estimated payment amount (amount and standard). Deviation) is calculated.
Area M2: Average amount 100 yen, standard deviation 30
Area M3: Average amount 8,000 yen, standard deviation 500

The total expected payment amount calculation unit 13 calculates the total expected payment amount added in all the stay areas for each user.
User ID 1, specified period 8 / 19-8 / 23
Area of stay M2, M3, M2
Addition of total expected payment amount / average value = 8,000 yen x 1 time + 100 yen x 2 times = 8,200 yen Addition of total expected payment amount / standard deviation = √ (500 ² x 1 time + 30 ² x 2 times) = √ 25 1800 = 502
Assuming the independence of the distribution, the variance of the payment amount is the sum of the variances of the places visited.

[User behavior change determination unit 14]
The user behavior change determination unit 14 determines for each user whether or not the total expected payment amount has changed to a predetermined amount or more according to the elapse of a predetermined period.
Here, the determination of "changed to a predetermined amount or more" is to determine whether or not the total expected payment amount in the previous predetermined period has changed to a predetermined amount or more than the total expected payment amount in the later predetermined period. May be.
Further, the user behavior change determination unit 14 separately determines whether or not the total expected payment amount has changed to a predetermined amount or more, or the total expected standard deviation has changed to a predetermined standard deviation or more.

FIG. 7 is an explanatory diagram showing the processing of the user behavior change determination unit.

According to FIG. 7, the user behavior change determination unit 14 makes a determination based on the two-sample t-test.
The "two-sample t-test" refers to testing whether or not there are two independent populations and there is a difference in the average of the samples extracted from each population (see, for example, Non-Patent Document 3). According to the present invention, the two populations mean a user whose behavior has changed and a user whose behavior has not changed.

(S1) The F-test of the ratio of the population variance determines whether the null hypothesis of homoscedasticity is rejected or rejected.
The variance ratio is calculated assuming that the variance of the past predetermined period Tn and the variance of the current predetermined period Tm follow the chi-square distribution of degrees of freedom n-1 and m-1, respectively. This dispersion ratio follows an F distribution with n-1 and m-1 degrees of freedom. When the significance level is α, the variance ratio is larger than F (n-1, m-1, α / 2) or smaller than F (n-1, m-1,1-α / 2), etc. The assumption of variance is rejected.

(S21) When the null hypothesis of homoscedasticity is not rejected (in the case of homoscedasticity), the presence or absence of a significant difference is determined by a t-test assuming that the variances are similar. Here, after calculating the invariant variance, t is tested.
s _n ² = Σ _{i = 1} ⁿ (xi ^- x _n- ) ² / (n-1)
sm ² = Σ _{i = 1 m} (xi ^－ x ^m- ) ² / ( _m -1)
s ² = {(n-1) x s _n ² + ( _m -1) x sm ² } / (n + m-2)
t = (x _n ^--- x _m- ⁾ / s√ (1 / n + 1 / m)
If it is determined by the t-test that there is a significant difference, the flag "1" is set as there is a change in user behavior.
On the other hand, if it is determined by the t-test that there is no significant difference, the flag "0" is set as no change in user behavior.

(S22) When the null hypothesis of homoscedasticity is rejected, the presence or absence of a significant difference is determined by Welch's t-test assuming that the variances are different.
t = (x _n ^--- x _m- ⁾ / s√ (1 / n + 1 / m)
The degree of freedom v tests t using the following approximation.
v = (s _n ² / n + sm ² / _m ) ² / {s _n ² / (n ² (n-1) + sm ² / ( _m ² (m-1)}
If it is determined by Welch's t-test that there is a significant difference, the flag "1" is set as there is a change in user behavior.
On the other hand, if it is determined by Welch's t-test that there is no significant difference, the flag "0" is set as no change in user behavior.

Then, the user behavior change determination unit 14 outputs a flag for each user to the user group behavior change determination unit 15.

[User group behavior change determination unit 15]
The user group behavior change determination unit 15 activates that a behavior change has occurred in the user group when the ratio of users determined to be true by the user behavior change determination unit 14 becomes a predetermined ratio or more.
Specifically, if a behavioral change occurs at a predetermined ratio of α% or more of the employee group of the company, some kind of alarm is given to the company.

As another embodiment, it is also preferable to determine that when a behavioral change occurs in the employee group of the target company, the behavioral change does not occur in the employee group of the other company. In particular, if a sudden deterioration in business conditions occurs in the country or the world, behavioral changes will occur not only in the target companies but also in the employee groups of other companies. That is, it is determined that the behavioral change of the employee group of the target company is not based on the sudden deterioration of business conditions.
In particular, when looking at a company as a whole, when there are a considerable number of companies in which behavioral changes have occurred in the employee group, the deterioration of business conditions has adversely affected the economic situation of that company, and the target company alone. It is different from the deterioration of the economic situation in Japan.

As described above in detail, according to the program, apparatus and method of the present invention, it is possible to determine the behavioral change of the user group.

According to the present invention, only the movement history in the employee group of the target company is an observation target. Behavioral changes in employee users are generally caused by a sharp drop in income. In particular, the number of visits to places that are leisure costs is expected to decrease significantly. Of course, because there are changes in the family structure and some users suddenly become eco-friendly, we observe behavioral changes as a group of employees as a whole. If the behavioral changes of the entire employee group occur at a predetermined rate or more, it can be estimated that the cause is the economic situation of the target company.

The target company can know in advance whether or not there is a change in the behavior of the employee group when the salary is reduced or the payment is delayed for the employee due to his / her business situation. Even before the target company announces its own business situation, it is possible to consider measures for employees.
From the perspective of the management of the target company, it also helps to protect the interests and rights of employees by picking up the voices of employees. You can also consider what measures you can take for your employees and what impacts they can have on which parts.

On the other hand, the banks, investors, and business partners of the target company can predict the deterioration of the economic condition of the target company in advance before the business condition of the target company is announced by the financial statements.
From the perspective of a bank, etc., it is extremely useful to know whether or not there is a change in the behavior of a group of employees that has not been announced in the financial statements regarding changes in the economic situation of the company, in determining the creditworthiness of the company. If the behavior of the employee group changes, it is possible to start negotiations with the target company in advance, such as strengthening maintenance, collecting products, and reviewing conditions.

According to the present invention, it is classified as one of the scoring models, but the financial information of the target company is not required. In addition, the elapsed time until the target company deteriorates in business conditions is effective for both long-term and short-term.

Various changes, modifications and omissions of the technical idea and the scope of view of the present invention can be easily carried out by those skilled in the art with respect to the various embodiments of the present invention described above. The above explanation is just an example and does not attempt to limit anything. The present invention is limited only to the scope of claims and their equivalents.

1 Estimator 11 User position database 12 Estimated payment amount database 13 Total expected payment amount calculation unit 14 User behavior change judgment unit 15 User group behavior change judgment unit 2 Mobile terminal

Claims (12)

It is a program that makes the computer installed in the device that judges the behavior change of the user group function.
A user position database that accumulates time-series positions on mobile terminals owned by each user in the user group, and
An estimated payment amount database that stores the estimated payment amount in association with each region,
For each user, for each predetermined period, the user location database is used to identify one or more areas of stay, and the estimated payment amount database is used to add up the estimated payment amounts of all areas of stay within the specified period. A means of calculating the total expected payment amount to calculate the amount, and
A user behavior change determining means for determining whether or not the total expected payment amount has changed to a predetermined amount or more according to the elapse of a predetermined period for each user.
When the ratio of users determined to be true by the user behavior change determination means exceeds a predetermined ratio, the computer functions as a user group behavior change determination means that activates that a behavior change has occurred in the user group. Characterized program. The user group is the employee group of the company,
The program according to claim 1, wherein the occurrence of a change in behavior of a group of users further causes a computer to function so as to mean the occurrence of a change in the economic situation of the company. The program according to claim 1 or 2, wherein the means for calculating the total expected payment amount is to make the computer function so as to extract only the area where the user has stayed for a predetermined time or longer as the area of stay. The user behavior change determination means causes the computer to function so as to determine whether or not the total expected payment amount in the previous predetermined period has changed to a predetermined amount or more than the total expected payment amount in the later predetermined period. The program according to any one of claims 1 to 3, which is characterized. The expected payment amount database associates the average amount and standard deviation as the expected payment amount for each region.
The total expected payment amount calculation means calculates, for each user, the total expected payment amount obtained by adding the average amount of the expected payment amount and the total expected standard deviation which is the square root of the added value of the squared (variance) of the standard deviation.
The user behavior change determining means is characterized in that the computer functions to determine whether or not the total expected payment amount has changed to a predetermined amount or more, or the total expected standard deviation has changed to a predetermined standard deviation or more. The program according to any one of claims 1 to 4. The user behavior change determination means is based on a two-sample t-test.
The F-test of the ratio of the population variance determines whether the null hypothesis of homoscedasticity is rejected or rejected.
If the null hypothesis of homoscedasticity is not rejected, the presence or absence of a significant difference is determined by the t-test assuming that the variances are similar.
The program according to claim 5, wherein when the null hypothesis of equal variance is rejected, the computer is made to function to determine the presence or absence of a significant difference by Welch's t-test assuming that the variances are different. .. With respect to the user position database, the time-series positions are such that the terminal positioning position is positioned by the mobile terminal possessed by the user and / or the base station positioning position based on the stay range of the mobile terminal connected to the base station. The program according to any one of claims 1 to 6, wherein the computer is operated. A program that activates a computer installed in a device that determines changes in a company's economic situation.
For each user who is an employee of the company, the total expected payment amount that adds up each expected payment amount and the total expected standard deviation that is the square root of the variance of the expected payment amount is calculated for each predetermined period. Calculation means and
As a user behavior change determining means for determining whether or not the total expected payment amount has changed to a predetermined amount or more or the total expected standard deviation has changed to a predetermined standard deviation or more according to the elapse of a predetermined period for each user. ,
When the ratio of users determined to be true by the user behavior change determination means exceeds a predetermined ratio, the computer functions as a user group behavior change determination means that activates that a behavior change has occurred in the user group. Characterized program. A device that determines behavioral changes in a group of users.
A user position database that accumulates time-series positions on mobile terminals owned by each user in the user group, and
An estimated payment amount database that stores the estimated payment amount in association with each region,
For each user, for each predetermined period, the user location database is used to identify one or more areas of stay, and the estimated payment amount database is used to add up the estimated payment amounts of all areas of stay within the specified period. A means of calculating the total expected payment amount to calculate the amount, and
A user behavior change determining means for determining whether or not the total expected payment amount has changed to a predetermined amount or more according to the elapse of a predetermined period for each user.
It is characterized by having a user group behavior change determining means that activates that a behavior change has occurred in the user group when the ratio of users determined to be true by the user behavior change determining means becomes a predetermined ratio or more. Equipment to do. It is a device that judges changes in the economic situation of a company.
For each user who is an employee of the company, the total expected payment amount that adds up each expected payment amount and the total expected standard deviation that is the square root of the variance of the expected payment amount is calculated for each predetermined period. Calculation means and
As a user behavior change determining means for determining whether or not the total expected payment amount has changed to a predetermined amount or more or the total expected standard deviation has changed to a predetermined standard deviation or more according to the elapse of a predetermined period for each user. ,
It is characterized by having a user group behavior change determining means that activates that a behavior change has occurred in the user group when the ratio of users determined to be true by the user behavior change determining means becomes a predetermined ratio or more. Equipment to do. It is a method for determining a change in behavior of a user group of a device.
The device is
A user position database that accumulates time-series positions on mobile terminals owned by each user in the user group, and
An estimated payment amount database that stores the estimated payment amount in association with each region,
Have,
The device is
For each user, for each predetermined period, the user location database is used to identify one or more areas of stay, and the estimated payment amount database is used to add up the estimated payment amounts of all areas of stay within the specified period. The first step to calculate the amount and
For each user, the second step of determining whether or not the total expected payment amount has changed to the predetermined amount or more according to the lapse of a predetermined period, and
A user characterized by executing a third step of invoking that a behavioral change of a user group has occurred when the ratio of users determined to be true by the second step exceeds a predetermined ratio. Group behavior change judgment method. It is a method for judging changes in the corporate economy of equipment.
The first step to calculate the total expected payment amount, which is the sum of the expected payment amounts, and the total expected standard deviation, which is the square root of the variance of the expected payment amount, for each user who is an employee of the company. When,
For each user, the second step of determining whether the total expected payment amount has changed to the predetermined amount or more or the total expected standard deviation has changed to the predetermined standard deviation or more according to the lapse of a predetermined period.
It is characterized by executing the third step of invoking that a change in the economic situation of a company has occurred when the ratio of users determined to be true by the second step exceeds a predetermined ratio. Corporate economic change judgment method.

Images