Game information management system

Abstract

To provide a game information management system for predicting the operating situation of a new model in a phase before its introduction.SOLUTION: A game information management system 1 for a game house where a plurality of models of game machines 2 are installed can collect players' posts about a new model on a prescribed SNS site for a prescribed period before and after the date of introduction of a game machine 2 of the new model as the subject period, and generate a learning model for extracting words included in the posts, determining the polarity of each word indicating at least whether it is positive or negative, and carrying out machine learning by using the polarity ratio per machine model in all the posts and the operating data per machine model as teacher data to predict the operating situation of game machines.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a game information management system. [Background technology]

[0002] Conventionally, a system has been proposed for predicting the operating status of a new machine model introduced in an amusement facility, as described in, for example, the following Patent Document 1. Such a system makes it possible to predict how long the introduced new machine model will be able to maintain good operating performance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2017-176586 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, the above-mentioned conventional systems can only make predictions based on actual operation data after installation, and therefore have the problem of being useless at the stage before installation.

[0005] The present invention has been made in consideration of the above-mentioned problems in the conventional art, and aims to provide a game information management system for predicting the operating status of a new machine type before it is introduced. [Means for solving the problem]

[0006] The gaming information management system of the present invention is a management system for an amusement parlor where multiple types of gaming machines are installed. This gaming information management system collects player posts on a specified SNS site regarding a new type of gaming machine during a specified period before and after the date on which the new type of gaming machine is introduced.

[0007] The gaming information management system of the present invention extracts words included in posts, determines the polarity of each word, indicating at least whether it is positive or negative, and calculates the polarity ratio of the entire posts as the polarity ratio by machine type. The gaming information management system then performs machine learning using the operation data and polarity ratio by machine type as training data, and can generate a learning model for predicting the operation status of the gaming machines.

[0008] According to the gaming information management system of the present invention, a learning model useful for predicting the operation status after the introduction of a new gaming machine model can be generated. By using this learning model, it becomes possible to predict the operation status after the introduction of a new gaming machine model before the introduction. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a system diagram showing the configuration of a game information management system. [Diagram 2] FIG. 2 is a front view of the pachinko game machine and the card unit. [Diagram 3] FIG. 2 is a block diagram showing the electrical configuration of the pachinko game machine and the card unit. [Figure 4] FIG. 4 is an explanatory diagram of polarity dictionary data managed by the management server. [Diagram 5] FIG. 4 is an explanatory diagram of collected post data managed by a management server. [Figure 6] FIG. 11 is an explanatory diagram of daily aggregate data of the number of posts managed by the management server. [Figure 7] FIG. 11 is an explanatory diagram of period-aggregated data of the number of posts managed by the management server. [Figure 8] 4 is an explanatory diagram of weekly average out data managed by the management server. [Figure 9] An explanatory diagram of model-specific data managed by the management server, including average polarity ratio, average number of posts, number of players, and operating lifespan. [Figure 10] FIG. 4 is an explanatory diagram of data for each gaming machine managed by the gaming facility management device. [Figure 11] FIG. 4 is an explanatory diagram of machine-specific data managed by the amusement facility management device. [Figure 12]FIG. 11 is an explanatory diagram of period transition data displayed by the amusement facility management device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] In the present invention, when a new type of gaming machine is introduced to an amusement facility, the polarity ratio of the new machine can be input into the learning model to predict the operating status after the introduction. By using this method, when introducing a new type of gaming machine, it is possible to plan the appropriate number of machines to be introduced and the increase or decrease of machines after the introduction at an earlier stage. This reduces the misjudgment when introducing a new machine, and realizes sound management of the amusement facility.

[0011] The embodiment of the present invention will be specifically described using the following examples. Example 1 This example is an example of a gaming information management system 1 for an arcade where multiple types of gaming machines 2 are installed. This gaming information management system 1 can provide information that supports the arcade's decision when introducing a new type of machine. The details will be described with reference to Figs. 1 to 12.

[0012] The gaming information management system 1 (FIG. 1) of this example is configured to include an in-game system 10S in an amusement facility where gaming machines 2 are installed, and an external management server 11. The in-game system 10S and the management server 11 are connected to a public communication line such as the Internet and are capable of communicating with each other. On the Internet, SNS sites 13 such as short message posting sites such as the old Twitter (registered trademark) and pachinko-related bulletin board sites are open to the public.

[0013] The management server 11 is an external server managed by the operating company of the gaming information providing site, the gaming machine manufacturer, etc. The management server 11 has a function of feeding back the aggregated data of gaming information and the machine evaluation information to each gaming facility. The management server 11 collects gaming data from each gaming facility and creates various aggregated data. Furthermore, the management server 11 in this example can create the machine evaluation information by social listening.

[0014] In each amusement parlor, a card unit (score processing device) 4 that provides the gaming value required for playing, a gaming information display device 105, and the like are individually installed for gaming machines 2 such as pachinko gaming machines 21 and pachislot gaming machines (not shown). Furthermore, one relay device 109 is installed for every two gaming machines 2. In the management space in the amusement parlor, an amusement parlor management device 10 is installed for managing various devices in the amusement parlor. In the prize exchange counter (not shown) in the amusement parlor, a prize exchange terminal (not shown) and the like that exchanges the gaming value acquired by the player for prizes are installed.

[0015] The gaming facility equipment such as the gaming facility management device 10, the relay device 109, the prize exchange terminal, etc. are communicatively connected to an in-flight network 100. The card unit 4 and the game information display device 105 are communicatively connected to the relay device 109. Furthermore, the gaming machine 2 is communicatively connected to the card unit 4 via a dedicated line.

[0016] In the in-house network 100, various communications between the gaming machines 2, including peripheral devices such as a gaming information display device 105, and the gaming facility management device 10 are realized via a relay device 109. In each gaming facility, an in-house system 10S is constructed using the in-house network 100.

[0017] The amusement facility management device 10 is connected to the in-facility network 100, and is also connected to the Internet, which is a public communication line. In the game information management system 1, the in-facility system 10S and the management server 11 outside the amusement facility are connected in a communicable state via the amusement facility management device 10 connected to the Internet.

[0018] The gaming center management device 10 manages the operating status of all the devices installed in the gaming center (such as the gaming machines 2, the card units 4, and the gaming information display device 105). The gaming center management device 10 compiles the gaming information collected from each gaming machine 2, calculates and displays various gaming data, and transmits the gaming data to the management server 11. The gaming center management device 10 can also display the compiled data of the gaming center received from the management server 11.

[0019] The relay device 109 is a communication device that mediates communications between the corresponding gaming machine 2, the card unit 4, the game information display device 105, and the gaming facility management device 10. As described above, the relay devices 109 are installed at a ratio of one for every two gaming machines.

[0020] The card unit (score processing device) 4 is installed for each gaming machine 2, and is a device that executes score-related processes such as the process of granting the scores required for playing and the process of issuing cards on which the scores are recorded. The card unit 4 also has the function of receiving various game information from the gaming machine 2 and transmitting it to the gaming facility management device 10 and the game information display device 105 via the relay device 109. In the explanation of this example, counted scores are referred to as "owned points", and scores that are available for playing are referred to as "game points".

[0021] The gaming information display device 105 is installed for each gaming machine 2, and displays various information related to the game of the corresponding gaming machine 2. The gaming information display device 105 receives various pieces of gaming information output by the gaming machine 2 via the card unit 4 and the relay device 109. The gaming information display device 105 creates and displays various pieces of display information based on the various pieces of gaming information received. The gaming information display device 105 is installed, for example, above the corresponding gaming machine 2.

[0022] In an amusement center, for example, about 300 to 1500 gaming machines 2 of multiple models, about 100 to 500 types, are installed. Each gaming machine 2 can play using awarded or acquired gaming points, and transmits various types of game information to a relay device 109 according to the execution of a game. As will be described in detail later, the gaming machine 2 outputs various types of information that differ depending on the model.

[0023] Next, (A) the configuration and basic operation of the gaming machine and card unit, (B) the configuration and management data of the management server, (C) the configuration and management data of the gaming center management device, and (D) the display data by the gaming center management device will be explained in that order.

[0024] (A) Configuration and basic operation of gaming machine and card unit A pachinko game machine 21 (FIG. 2), which is an example of the game machine 2, is a game machine in which a game is played by consuming (using) game points (scores) that are game value to shoot game balls that are game media. The illustrated pachinko game machine 21 is a so-called seven-machine that executes a jackpot lottery in response to winning in the special chart start port 221, and displays a jackpot pattern when a jackpot is won, generating a jackpot state.

[0025] As shown in Fig. 2, the pachinko game machine 21 has a substantially circular play area (play board) 210. Decorative lamp units 242 are provided on both the left and right sides of the upper part of the play area 210, and a speaker 241 is disposed on the lower left side of the play area 210. In addition, an information display unit 25 is disposed below the play area 210, and an operating handle 26 is provided on the lower right of the information display unit 25.

[0026] The information display unit 25 is composed of a liquid crystal display, and various information such as various game information related to the game, such as the number of jackpots and the number of special drawing draws, is displayed on the display screen of the information display unit 25. The display screen of the information display unit 25 is a display screen in which a sheet-like touch panel 251 (FIG. 3) is laminated and which can be operated by touch. The display contents of the information display unit 25 can be switched according to the touch operation on the display screen.

[0027] The display screen of the information display unit 25 displays the accumulated game points, such as the game points received from the card unit 4 and the game points based on the points awarded in response to winning. The game points are subtracted by one each time a game ball is shot, while a predetermined game point (score) is added in response to the winning of a game ball into a winning slot. In the pachinko game machine 21, when the game points reach zero, it becomes impossible to shoot game balls.

[0028] The game area 210 is an area where game balls, which are game media, flow down. In the game area 210, a display device including a liquid crystal display unit 24 is arranged near the center, and an outlet hole 225 for collecting game balls that flow down without winning is provided at the bottom. The special chart start port 221 is arranged below the liquid crystal display unit 24, and a large winning port 222 is provided between the special chart start port 221 and the outlet hole 225.

[0029] On the left side of the liquid crystal display unit 24 as viewed from the player's side, a pass-gate-type regular symbol start port 223 and the like are arranged. The regular symbol start port 223 is a gate for detecting the passage of a gaming ball. The regular symbol start port 223, which only allows gaming balls to pass, is not set for adding gaming points. When a gaming ball is detected by the regular symbol start port 223, a lottery random number for determining whether or not the regular symbol is a winner (hereinafter referred to as the regular symbol lottery) is extracted, and the regular symbol lottery is executed. The result of the regular symbol lottery is displayed by the liquid crystal display unit 24.

[0030] The special symbol start port 221 is equipped with an electric tulip (electric role), and is a variable winning port that triggers the judgment of whether the special symbol is a winning symbol (hereinafter referred to as the special symbol lottery). The electric tulip opens when a winning symbol is selected in the normal symbol lottery, and significantly increases the winning rate of the special symbol start port 221. When a game ball enters the special symbol start port 221 (start winning), a lottery random number for the special symbol lottery is extracted, and the special symbol lottery (jackpot lottery) is executed, and a predetermined number of game points (for example, 3 points) are added.

[0031] The liquid crystal display unit 24 is a display unit that variably displays the patterns using a liquid crystal display. The liquid crystal display unit 24 has a function as a special pattern display unit that displays the result of the special pattern lottery (lottery result) using patterns, and a function as a normal pattern display unit that displays the result of the normal pattern lottery (lottery result). If the special pattern lottery is won, the win is notified by the liquid crystal display unit 24 as a special pattern display unit, and then a jackpot occurs, including the opening of the jackpot opening 222. If the jackpot opening 222 is won, a predetermined number of game points (for example, 15 points) are added.

[0032] A pachinko gaming machine 21 is configured around a main control unit 20 as shown in Fig. 3. The main control unit 20 includes a CPU (Central Processing Unit) 201, a ROM (Read Only Memory) 202 and a RAM (Random Access Memory) 204 which are memory elements, and an I / O (Input / Output) 205 which serves as an input / output interface.

[0033] The memory area of ​​the RAM 204 is provided with memory areas for storing game points, which indicate the number of game balls that can be shot, game points, which are the current value of the game value acquired by the player, etc. The main control unit 20 including the RAM 204 has a function as a game value storage means, etc.

[0034] The main control unit 20 is electrically connected to an information display unit 25, a touch panel 251, a performance control unit 291, winning game ball sensors 293 to 295, a special prize start solenoid 296 that opens the special prize start port (electric chute) 221, a special prize opening solenoid 297 that opens the special prize opening 222, a power supply circuit unit 298 for supplying power, a setting operation unit 23, etc. Furthermore, an I / F unit 200 that serves as a communication window with the card unit 4 is electrically connected to the main control unit 20.

[0035] The performance control unit 291 is electrically connected to an amplifier 241A that drives the speaker 241 and a decorative lamp unit 242, and is also communicatively connected to a display control unit 292 that controls the display screen of the liquid crystal display unit 24.

[0036] Sensors for detecting winning game balls include a special prize sensor 293 that detects prizes in the special prize start port 221, a regular prize sensor 294 that detects prizes in the regular prize start port 223, and a special prize sensor 295 that detects prizes in the special prize port 222.

[0037] The setting operation unit 23 is an operation unit for performing various settings. The setting operation unit 23 is provided inside the gaming machine main body, and can be operated when the gaming board surface is opened forward. The setting operation unit 23 can perform a setting operation to set one of, for example, six setting values ​​with different jackpot probabilities.

[0038] The pachinko gaming machine 21 transmits various types of gaming information to the card unit 4 in response to the execution of a game in which gaming points are consumed. The gaming information output by the gaming machine 2 includes various types of information that differ depending on the model. For example, the pachinko gaming machine 21 in this example transmits the following gaming information to the card unit 4. The gaming information output by the pachinko gaming machine 21 is transmitted to the relay device 109 via the card unit 4, and is further transmitted from the relay device 109 to the gaming facility management device 10, the gaming information display device 105, etc.

[0039] · Information on number of game points consumed: Information indicating the number of game points consumed in a game. Awarded game points information: Information indicating the number of game points awarded upon winning. Start information: Information indicating that a jackpot lottery (change in special chart) has been held. · Jackpot information: Information indicating that a jackpot has occurred. - Chance bonus information: Information indicating that the game is in a chance bonus state.

[0040] As described above, the card unit 4 is installed for each gaming machine 2, and is a device that performs processes related to scores, such as the process of granting points required for playing and the process of issuing cards on which points are recorded. The card unit 4 is installed, for example, in a vertically long gap between the corresponding gaming machine 2 and the adjacent gaming machine 2. The card unit 4 transmits various types of game information received from the gaming machine 2 to the gaming facility management device 10 and the game information display device 105 via the relay device 109.

[0041] The card unit 4 executes a process of granting points required for playing games by using a granting function (lending function) of game points (scores) that are the right to launch game balls, and also executes processes related to the scores, such as issuing a game card on which the scores are recorded. The card unit 4 is an example of a providing device that provides a predetermined service for members, such as a ball saving service, on the condition that a member ID, which is identification information of a member registered with the game facility, is authenticated.

[0042] On the front of the card unit 4 (see Figure 2), there are arranged a status lamp 41 which indicates the operating status of the device such as errors, a bill insertion slot 420, a touch-operable display unit 43, a loan button 44 for receiving a loan of game points in exchange for a withdrawal of payment, a re-play button 45 for converting the counted points into game points, a return button 46 for requesting the return of the game card, and a card insertion slot 470 for the game card.

[0043] 3, the card unit 4 is electrically configured with a control unit 40 including a CPU 401, a ROM 402, a RAM 403, an I / O 404, etc., as a central component. In addition to the above components, a banknote processing unit 42, a card processing unit 47, and an I / F unit 400 are electrically connected to the control unit 40. The card unit 4 is capable of communicating with the relay device 109 via the I / F unit 400, and is also directly connected to the corresponding pachinko gaming machine 21.

[0044] Next, the basic operation of the pachinko game machine 21 and the card unit 4 will be described. When the pachinko game machine 21 receives game points (scores) from the corresponding card unit 4, it becomes able to launch game balls equal to the game points. The pachinko game machine 21 subtracts one game point from the game points every time a game ball is launched, while adding points according to the winning to the game points when a prize is won. When the pachinko game machine 21 reaches zero, it becomes impossible to launch game balls and play cannot be continued.

[0045] In the pachinko game machine 21, a game ball can be launched from a launching device (not shown) located at the lower right of the game area 210 by operating the operation handle 26. When the game ball enters the special chart start hole 221, a special chart lottery (jackpot lottery) is executed, and the special chart pattern on the liquid crystal display unit 24 starts to change.

[0046] If the special drawing is won, the liquid crystal display 24 as the special drawing display unit displays a jackpot symbol such as 777 to notify the jackpot, and the state transitions to the jackpot state. In the jackpot state, the jackpot winning slot 222 is opened a specified number of times, greatly improving the winning rate. If the special drawing is not won, the liquid crystal display 24 displays a predetermined losing symbol.

[0047] In addition, when the regular lottery is won at the regular lottery start port 223, a regular lottery is executed, and the pattern variation starts on the liquid crystal display unit 24 as a regular lottery display unit. When the regular lottery is won, the liquid crystal display unit 24 notifies the fact, and the electric tulip of the special lottery start port 221 opens, making it easier to win at the special lottery start port 221.

[0048] The gaming states of the pachinko gaming machine 21 include a jackpot state, a normal state, and a probability variable state. The probability variable state is a hot state in which the probability of winning the special symbol judgment is higher than in the normal state, and the expectation of a jackpot is high. In the illustrated pachinko gaming machine 21, after the jackpot state ends, there is a 70% probability of transitioning to the probability variable state, and a 30% probability of transitioning to the normal state. Note that performance and playability, such as the probability of winning a jackpot, probability variable state, or time-saving state, the number of points awarded when winning, and the manner of pattern change, differ depending on the model of the gaming machine 2.

[0049] When the card unit 4 accepts a bill, it adds the amount of the bill to the deposit balance (up to 10,000 yen) and stores it. When a gaming card is accepted, it matches the card ID, which is the identification information of the gaming card, with the card information stored in the gaming facility management device 10. Then, when the matching result is OK, the card unit 4 displays the deposit balance and the number of points related to the card information.

[0050] The card unit 4 converts any of the player's deposit balance, saved balls (points acquired and accumulated up to the previous day), and owned points into game points in response to the player's operation, and transmits them to the corresponding pachinko gaming machine 21. For example, when the lending button 44 is operated, the card unit 4 converts a fixed amount (e.g., 1000 yen) into game points (250 points) at a time, and transmits them to the pachinko gaming machine 21. For example, when the replay button 45 is operated, the card unit 4 converts the owned points into game points of 250 points at a time, and transmits them to the pachinko gaming machine 21, enabling replay with the owned points (replay processing).

[0051] The card unit 4 stores and displays points such as the number of points owned, the number of game points, and the total number of points which is the sum of the number of points owned and the number of game points. The card unit 4 receives point addition and subtraction information from the corresponding pachinko gaming machine 21 and updates the stored number of points.

[0052] When finishing playing on the pachinko gaming machine 21, it is necessary to have the number of game points stored in the pachinko gaming machine 21 counted. To have the number of game points counted, a counting operation (counting operation) of a counting button (not shown) displayed on the information display unit 25 is performed. The game points are transmitted to the card unit 4 in response to the counting operation, and the game points on the gaming machine side are subtracted. A short touch counting operation transmits game points one by one, and a long touch (continuous touch) counting operation transmits all game points at once. The card unit 4 converts the game points received from the pachinko gaming machine 21 into owned points. The game points stored in the pachinko gaming machine 21 become zero when transmission to the card unit 4 is completed.

[0053] When the return button 46 is operated while there is at least one of the deposited balance and the owned points, the card unit 4 records the deposited balance and the owned points on a gaming card and issues it. The gaming points stored in the pachinko gaming machine 21 are not recorded on the gaming card, but remain stored on the gaming machine side. The card unit 4 has a built-in card stocker, and can store up to 10 gaming cards.

[0054] The card unit 4 receives various types of game information from the corresponding pachinko gaming machine 21, and collects and stores, for example, the following operation data. (1) Number of jackpots: The number of times a jackpot state occurred. (2) Inter-jackpot start: The number of starts, which is the number of jackpot draws since the previous jackpot state ended (the number of current starts). (3) Today's cumulative starts: The cumulative number of starts since opening on that day. (4) Points consumed: The number of game points consumed by the player for playing. (5) Points awarded: Gaming points awarded to players depending on their winnings.

[0055] The card unit 4 transmits operation data to the gaming center management device 10 and the game information display device 105 at any time via the relay device 109. Furthermore, the card unit 4 also transmits information on the player's possession value, such as the player's deposit balance and points held, to the gaming center management device 10 at any time. The card unit 4 also receives and stores various setting information (gaming machine number, model name, point unit price, etc.) from the gaming center management device 10.

[0056] (B) Configuration and management data of the management server The management server 11 is a server device that stores aggregated data based on game data collected from each gaming facility via the Internet. The management server 11 has a function of collecting and analyzing information posted by players on social media by accessing SNS sites 13 on the Internet, such as short message posting sites such as the former Twitter (registered trademark) and pachinko-related bulletin board sites.

[0057] The management server 11 has the functions as the following means. (1) Post collection means: means for collecting player posts on a predetermined SNS site 13. In particular, the post collection means of this example is capable of collecting player posts regarding a new model of gaming machine 2 during a predetermined period around the date on which the new model of gaming machine 2 is introduced. The target period is, for example, a 24-week period from 8 weeks before the model is introduced to 16 weeks after the introduction. (2) Polarity determination means: A means for extracting words contained in the posts collected by the post collection means and determining the polarity, indicating at least positive or negative, for each word by model. (3) Polarity ratio calculation means: A means for calculating the polarity ratio of the entire posts as a polarity ratio for each model based on the result of the determination by the polarity determination means. The polarity ratio calculation means calculates the polarity ratio every time a certain number of days has passed in the target period. In this example, the certain number of days is one week. (4) Post number calculation means: means for calculating the number of posts collected by the post collection means. The post number calculation means in this example calculates the number of posts every time a certain number of days elapses during the target period. (5) Operation data calculation means: a means for calculating operation data showing the actual operation status after a new type of gaming machine is introduced. (6) Polarity adjustment means: A means for performing machine learning using the polarity ratio and the operational data as training data, and adjusting the criteria for judgment by the polarity judgment means. (7) Learning model generation means: a means for generating a learning model for predicting the operation status of a gaming machine. The learning model generation means performs machine learning using the polarity ratio and operation data for each model, and the number of posts and operation data for each model as training data, to generate a learning model. (8) Data storage means: means for storing various data (FIGS. 4 to 9).

[0058] The data stored by the management server 11 as a data storage means includes polarity dictionary data (Figure 4), post collection data (Figure 5), daily aggregate data on the number of posts (Figure 6), period aggregate data on the number of posts (Figure 7), weekly average out data by model (Figure 8), model-specific data on average polarity ratio and average number of posts, number of players and operating lifespan (Figure 9), etc.

[0059] The polarity dictionary data in Figure 4 is dictionary data that is based on existing polarity dictionary data, with words specific to the pachinko industry added and the polarity adjusted. An example of an existing polarity dictionary that can be used is the "Japanese Evaluation Polarity Dictionary." In the polarity dictionary data in Figure 4, words contained in articles introducing machines that have been introduced in the past have been additionally learned as pachinko terms. The words for additional learning are labeled as positive, negative, or neutral before being applied to learning. It is also good to collect words for learning from information posted on the Internet, etc.

[0060] In the polarity dictionary data, a polarity label and a polarity value (number in Figure 4) are set for each word. Examples of polarity labels include negative, positive, and neutral. The default polarity values ​​are "-1" for negative, "1" for positive, and "0" for neutral. Words that do not exist in the existing dictionary or words with different meanings can be manually registered and adjusted. For example, "super hot" is a positive word that indicates a state in which there is high expectation of a jackpot occurring, but since it is not in the existing dictionary, it is a good idea to register it manually.

[0061] In addition, in order to improve the accuracy of the polarity dictionary data, the management server 11 of this embodiment performs machine learning using the polarity ratio and the actual operation data as teacher data, and can adjust the criteria for judgment by the polarity judgment means to fit the evaluation value and the operation data. According to such machine learning, theoretically, it is possible to derive the conclusion that if the evaluation value is high, the operational life is long, and if the evaluation value is low, the operational life is short. If the operational life is short despite the high evaluation value, it is advisable to adjust the polarity dictionary data so that the evaluation value is lowered.

[0062] The collected post data in Fig. 5 is a collection of online posts sorted by model. The management server 11 collects online posts by social listening, for example, from short message posting sites such as the old Twitter (registered trademark) (referred to as SNS1 in the figure) and bulletin board sites (referred to as SNS2 in the figure). In the configuration of this example, the collection period is set to a total of 24 weeks, from 8 weeks before the introduction of the model to 16 weeks after the introduction.

[0063] The management server 11 collects evaluation information on gaming machines (models) on the Internet, for example, by scraping. The management server 11 acquires old Twitter (registered trademark) tweet information on gaming machines (models) by, for example, old Twitter API (registered trademark, Application Programming Interface). In addition, the management server 11 acquires posted information on gaming machines (models) from pachinko-related bulletin board sites.

[0064] The management server 11 determines the polarity (negative, positive, neutral) of the post content and records it together with the polarity value. The management server 11 compares it with the polarity dictionary data (FIG. 4) to determine the polarity of each word. The management server 11 then uses natural language processing to determine the polarity of the entire post based on the polarity and context of each word included in the post.

[0065] As for natural language processing, for example, "BERT", a natural language processing model from Google, can be used. "BERT" can determine not only the meaning of individual words but also the context, making it suitable for analyzing the emotions of people such as players. Furthermore, "BERT" has the excellent characteristic of being applicable to various tasks without modifying the model structure. "BERT" can improve the accuracy of natural language processing by applying additional learning to existing models.

[0066] Additional learning is the learning of words specific to the pachinko industry, such as pachinko terms and words that describe games. Words are extracted and collected from texts such as past reviews of gaming machines. The words for additional learning are classified and labeled in advance as positive, negative, or neutral. By performing this additional learning, called fine tuning in "BERT", it becomes possible to perform highly accurate natural language processing. Fine tuning is a technique called transfer learning, which is performed to improve accuracy by utilizing a trained model with data outside the target task. When the additional learning words are applied to the trained model, the parameters of the trained model are updated and a model specialized for the target task can be generated.

[0067] The image of natural language processing in this example is that, for example, based on a post such as "Pachinko XX is promising. The first hit seems tough, but if you do get lucky, the shot is big. The normal performance is decent too," the average polarity value of the words included in the post is calculated. For example, the combination of the words "expectation" and "can" has a polarity value of "1," the combination of the words "first hit" and "hard" has a polarity value of "-1," the combination of the words "shot" and "big" has a polarity value of "1," and the combination of the words "performance" and "decent" has a polarity value of "1." Therefore, the average polarity value is (1-1+1+1)÷4=0.5.

[0068] The management server 11 can calculate the polarity ratio for each model. The management server 11 calculates the ratio of the number of positive posts to the total number of posts as the polarity ratio (positive ratio, positive ratio) for each week, for example. For example, if the total number of posts is 1000 and the number of positive posts is 750, the polarity ratio is 75%.

[0069] The daily aggregated data of the number of posts in Figure 6 is the number of posts by SNS for each day recorded by machine type. According to this daily aggregated data of the number of posts, it can be determined that the more posts there are, the higher the interest of players.

[0070] The period-based aggregated data on the number of posts in Figure 7 is based on the daily aggregated data (Figure 6), and records the number of posts by SNS, divided into those before and after the introduction of the model. Note that in order to perform a polarity analysis on a specific model, at least 400 to 500 posts are required. For No. 3 "Pachinko △△" in the figure, polarity analysis could not be performed due to an insufficient number of posts.

[0071] The weekly average out data by model in Figure 8 is data recorded by model of the weekly average out (average number of gaming points consumed per business day in a week) each week (an example of a fixed period) that has passed since the introduction of a new model. The management server 11 determines that the operational life of a model has expired when the weekly average out falls below a reference value, which is the weekly average out for all models, and ends recording of the weekly average out data for that model. Note that any fixed value (e.g., 20,000) may be set as the reference value for determining the operational life. The recording period is not limited to every week, and may be every specified number of days, such as every 5 days or every 10 days.

[0072] The model-specific data in Figure 9 is data recorded by model as training data for machine learning, including the average polarity ratio and average number of posts, the number of players in the week of introduction, and the operating lifespan. The average polarity ratio is the average polarity ratio in the target period. The average number of posts is the average number of posts in the target period. The operating lifespan is the lifespan that is determined when the model's weekly average out falls below the weekly average outs of all models. The operating lifespan represents the operating lifespan until the model's weekly average outs falls below the weekly average outs of all models. The number of players per machine (number of players / machine) is the number of players divided by the number of machines installed for that model.

[0073] In this example, two target periods for the average polarity ratio and number of posts are set: a period of 8 to 6 weeks before introduction, and a period of 1 week before introduction to 1 week after introduction. Average polarity ratio 1 and average number of posts 1 are data for the target period of 8 to 6 weeks before introduction. Average polarity ratio 1 and average number of posts 1 are used as training data for predicting the number of players. Average polarity ratio 2 and average number of posts 2 are data for the target period of 1 week before introduction to 1 week after introduction. Average polarity ratio 2 and average number of posts 2 are used as training data for predicting the operational lifespan.

[0074] The management server 11 generates a learning model for predicting the number of players and the operational lifespan. There are three patterns for generating a learning model to predict the number of players in the launch week: A pattern in which machine learning is performed using the average polarity ratio from 6 to 8 weeks prior to launch and the number of players in the actual launch week as training data to generate a learning model. A pattern in which machine learning is performed using the average number of posts from 6 to 8 weeks prior to launch and the number of players in the week of actual launch as training data to generate a learning model. A pattern in which machine learning is performed using the average polarity ratio and average number of posts from 6 to 8 weeks prior to launch, as well as the number of players in the actual launch week, as training data to generate a learning model.

[0075] There are three patterns for generating a learning model for predicting operational life: A pattern in which machine learning is performed using the average polarity ratio from one week before to one week after installation and the actual operating lifespan as training data to generate a learning model. A pattern in which machine learning is performed using the average number of posts from one week before to one week after implementation and the actual operating lifespan as training data to generate a learning model. A pattern in which machine learning is performed using the average polarity ratio and average number of posts from one week before to one week after implementation, as well as the actual operational lifespan, as training data to generate a learning model.

[0076] In this example, the number of players and operating lifespan are exemplified as the operation data to be predicted, but the operation data to be predicted is not limited to these. For example, a learning model can be generated by using various operation data such as the cumulative gross profit amount one month after installation and the average out as the operation data to be predicted.

[0077] (C) Configuration and management data of the amusement facility management device As described above, the amusement facility management device 10 (FIG. 1) is a computer device for managing the operating status of various machines in the amusement facility. The amusement facility management device 10 is equipped with interfaces such as input devices such as a keyboard and a mouse (not shown) and output devices such as a display 101 and a printer (not shown), and can be operated by a manager of the amusement facility.

[0078] The amusement facility management device 10 is connected to an in-facility network 100 and can compile game data for each gaming machine 2 installed in the amusement facility. The amusement facility management device 10 can exchange various information with the management server 11 via the Internet, which is a public communication line. Information that the amusement facility management device 10 transmits to the management server 11 includes, for example, game data for each gaming machine 2. Information that the amusement facility management device 10 receives from the management server 11 includes, for example, compiled data for the amusement facility, evaluation information for each model obtained through social listening, and the like.

[0079] The game center management device 10 has functions as the following means. (1) Operation data collection means: means for collecting operation data indicating the operation status of the gaming machine 2. In particular, the operation data collected by the operation data collection means in this example includes operation data indicating the operation status of a new model of gaming machine 2. (2) Operational data storage means: means for storing the operational data collected by the operational data collection means. (3) Display means: means for displaying the polarity ratio (e.g., the positive ratio) and the number of posts together with the operational data. The display mode may be a display on the display 101 or a printed display on a printer. The display means can display the polarity ratio and the number of posts in chronological order (described later with reference to FIG. 12).

[0080] The gaming facility management device 10 stores and manages various types of operation data. In particular, the gaming facility management device 10 of this embodiment stores and manages gaming machine-specific data (FIG. 10), which is operation data for each gaming machine, and model-specific data (FIG. 11), which is operation data for each model.

[0081] The data by gaming machine in Fig. 10 is operation data by gaming machine that is compiled and stored based on game information received from the gaming machine 2. In the data by gaming machine in the same figure, various game data is stored by machine number for each business day. The gaming facility management device 10 transmits the data by gaming machine to the management server 11 every time a predetermined time, such as every 10 minutes, has elapsed.

[0082] The items in the data for each gaming machine are as follows. · Model ID: Identification information (ID) to identify the model. · Points consumed: The number of points consumed in the game. It is equal to the number of balls fired. · Awarded points: The points awarded for winning. Start: The number of times the special chart changes in normal mode (the number of times the jackpot is drawn). · Jackpot: The number of times a jackpot has occurred. Number of players: The ID of the card used is used to determine the change of player and to count the number of players. It is also possible to attach a camera to each card unit 4 to capture an image of the player and count the number of players by image recognition.

[0083] The machine-specific data in FIG. 11 is data in which various game data is divided into each business day, compiled and stored by machine type. The payout rate is the average number of points awarded / average number of points consumed. The game time is the time during which consumed points or pattern changes occur continuously within a certain period of time. When consumed points or pattern changes occur again within, for example, two minutes after the occurrence of consumed points or pattern changes, it is determined that the game is continuing.

[0084] (D) Display data by the amusement facility management device The amusement facility management device 10 can display evaluation information for each model obtained by social listening together with the operation data for each model in Fig. 11. The evaluation information for each model is information received from the management server 11, and includes, for example, the number of posts related to the model, the polarity ratio in all posts, and the like.

[0085] The gaming facility management device 10 can, for example, display the time-series data of Fig. 12 on the display 101 or print it out. The time-series data is time-series data for each model on a weekly basis. This time-series data is useful data that helps the manager of the gaming facility to evaluate the models of the gaming machines 2.

[0086] The time period transition data is displayed in a horizontally long display field, with a model name display field 10A, a current (one week after introduction) data display field 10B, and a 10-week weekly data display field 10C covering the 10 weeks from 8 weeks before introduction to the week after the introduction. The current weekly data one week after introduction is displayed in the current (one week after introduction) data display field 10B, and also in the 10-week weekly data display field 10C. The total number of posts in the data display field 10B is the cumulative number of posts for the above 10 weeks.

[0087] In the 10 weeks' worth of weekly data display field 10C, the weekly data for the past 10 weeks is arranged (displayed) in chronological order, starting from the weekly data from 8 weeks ago on the left side. The right end of the 10 weeks' worth of weekly data display field 10C is the weekly data from one week after introduction, which corresponds to the present, and to the left of that is the weekly data from the introduction week, which corresponds to one week ago.

[0088] The composition of the weekly data in the display area 10C for 10 weeks of weekly data differs before and after the introduction. The weekly data before the introduction is composed of the positive ratio (in the figure, positive ratio; polarity ratio) and the number of SNS posts. The weekly data after the introduction includes the positive ratio and the number of SNS posts as well as out (number of gaming points consumed) and used price. Out is the average out for the relevant model over one business day, calculated based on data received from the gaming center management device 10 of each gaming center. The used price is price data received from a server (not shown) of a dealer that manages the distribution of used gaming machines.

[0089] The display format of the time-period transition data is not limited to this example, and various formats can be adopted. For example, the transition of each data may be displayed as a bar graph or a line graph. As data showing the operation status other than out, the time-period transition data may display, for example, the number of players, the playing time, the number of pattern changes, the number of jackpots, etc. In the example of the time-period transition data, the data is displayed in weekly intervals, but it is sufficient that the unit is a certain number of days, and the interval may be any number of days, such as every 10 days or every 5 days.

[0090] According to the period transition data illustrated in FIG. 12, the manager of the game center or the like can make the following judgments, for example. Since orders for models usually need to be placed at least five weeks in advance, the administrator, etc. can refer to data from six to eight weeks prior to the date of the order to determine whether to place an order and how many units to order.

[0091] Managers, etc. can decide whether to increase or decrease the number of machines based on data from the week of introduction to about one week after introduction. In addition, when the positive ratio and the number of posts change dramatically compared to before introduction, it is possible to decide that it is necessary to increase or decrease the number of machines. In the case of the time transition data in Figure 12, the manager of the game center can determine that although the positive ratio and the number of posts increase dramatically in the week of introduction and one week after introduction, the increase in the used price is gradual, so it is desirable to purchase used machines early and increase the number of machines. In addition, when the positive ratio drops dramatically after introduction, the manager can consider that the expectations of players before the introduction have been betrayed, and can consider selling the used machines early before the used price plummets.

[0092] The gaming information management system 1 of this example is a system that manages gaming information of an amusement center where multiple types of gaming machines are installed. This gaming center management system 1 extracts words included in posts, determines the polarity of each word, which indicates at least whether it is positive or negative, and calculates the polarity ratio of all posts as the polarity ratio by type.

[0093] In the gaming center management system 1 of this example, posts by players on a specific SNS site regarding a new gaming machine are collected and the number of posts is calculated for a specific period before and after the date on which the gaming machine is introduced to the gaming center as a new model. In addition, in this gaming information management system 1, operation data showing the operation status after the gaming machine is introduced as a new model is calculated.

[0094] The gaming information management system 1 can perform machine learning using the operation data and polarity ratio for each type of gaming machine, the number of posts and operation data for each type of gaming machine as teacher data, and generate a learning model for predicting the operation status of the gaming machine. The learning model generated by the gaming information management system 1 of this example is useful for predicting the operation status after the introduction of a new type of gaming machine. By using this learning model, it is possible to predict the operation status after the introduction of a new type of gaming machine at the pre-introduction stage. If the operation can be predicted before the introduction of a new type of gaming machine in this way, there is an advantage in that it is possible to plan the number of machines to be introduced and the increase or decrease in the number of machines after introduction at an early stage.

[0095] The gaming information management system 1 collects player posts on a specified SNS site regarding a new gaming machine 2 of a new model for a specified period around the date of introduction of the new gaming machine 2, calculates the number of posts, and determines the polarity of the posts, indicating whether the content of the posts is positive or negative. The gaming information management system 1 calculates the polarity ratio of all posts regarding the new gaming machine as a polarity ratio. The gaming information management system 1 then displays the calculated polarity ratio and the number of posts together with operation data indicating the operation status of the new gaming machine 2.

[0096] The management server 11, which serves as a polarity ratio calculation means, calculates the polarity ratio (positive ratio) every time a certain number of days elapses during the target period. The amusement facility management device 10, which serves as a display means, displays the polarity ratio and the number of posts in chronological order. Displaying the polarity ratio and the number of posts in chronological order in this manner allows the manager of the amusement facility, etc., to easily grasp changes in the polarity ratio and the number of posts. Note that displaying the polarity ratio and the number of posts in parallel is not a required configuration, and it is also possible to display either one in chronological order.

[0097] The management server 11 as the polarity adjustment means performs machine learning using the polarity ratio (positive ratio) and the operation data as teacher data, and adjusts the criteria for judgment by the management server 11 as the polarity determination means. In this way, machine learning using the polarity ratio and the operation data as teacher data can improve the accuracy of the polarity judgment by the management server 11 as the polarity determination means.

[0098] Instead of or in addition to the configuration of this embodiment, the following configurations may be adopted. Also, the following configurations may be adopted in appropriate combination. In this example, the predetermined period for collecting posts is set to 24 weeks, from 8 weeks before the model was introduced to 16 weeks after the introduction, but the target period is not limited to this example and may be shortened or extended as appropriate. Also, in this example, one week is given as an example of the fixed period, and the period until the weekly average out falls below the reference value, which is the weekly average out for all models, is set as the operational lifespan. The fixed period can be set to any period, such as 3 days, 10 days, etc., instead of the one week in this example.

[0099] In this example, the operating life is determined based on the average weekly outage of the model, but the operating life may be determined based on other operating data such as the operating rate and operating time, depending on the level of operation. The data by gaming machine and data by model shown in this example are merely examples, and various types of gaming information can be calculated and created. For example, the difference in points indicating the difference between the awarded points and the consumed points, the jackpot probability indicating the ratio of the number of jackpots to the number of starts, etc. may be calculated.

[0100] The gaming machine 2 illustrated in this example is merely one example, and the gaming information management system 1 can be configured for gaming machines of various specifications. For example, in this example, an enclosed type pachinko gaming machine that awards gaming points according to winning is illustrated, but the gaming machine is not limited to this. It may be a pachinko gaming machine that pays out gaming balls according to winning. It may also be a pachislot gaming machine (slot machine) or a jump ball gaming machine. The gaming value used in the game may be physical balls, medals, etc., in addition to electronic points.

[0101] It is also good to add a prediction means for predicting the operating status after the introduction of a new model. The prediction means is a means for predicting the operating status such as the number of players after the introduction of a new model and the operating lifespan by using the learning model generated by the learning model generation means. For example, if any of the average polarity ratio 8 weeks to 6 weeks before the introduction, the average number of posts 8 weeks to 6 weeks before the introduction, and the average polarity ratio and average number of posts 8 weeks to 6 weeks before the introduction are input into a learning model for predicting the number of players in the introduction week, the number of players in the introduction week can be predicted with high accuracy. Also, for example, if any of the average polarity ratio 1 week before the introduction to 1 week after the introduction, the average number of posts 1 week before the introduction to 1 week after the introduction, and the average polarity ratio and average number of posts 1 week before the introduction to 1 week after the introduction are input into a learning model for predicting the operating lifespan, the actual operating lifespan can be predicted with high accuracy.

[0102] Although specific examples of the present invention have been described in detail as examples, these specific examples merely disclose examples of the technology encompassed by the claims. Needless to say, the claims should not be interpreted as being limited by the configurations, numerical values, etc. of the specific examples. The claims encompass technologies that are obtained by variously modifying, changing, or appropriately combining the specific examples by utilizing publicly known technologies and the knowledge of those skilled in the art. [Explanation of symbols]

[0103] 1 Gaming Information Management System 10S In-house system 10 Game center management device (operation data collection means, operation data storage means, display means) 100 In-house network 101 Display 105 Gaming information display device 109 Repeater 11 Management server (post collection means, polarity determination means, polarity ratio calculation means, post number calculation means, operational data calculation means, polarity adjustment means, learning model generation means, data storage means) 13 Social Networking Sites 2. Gaming machines 21 Pachinko machines 4. Card unit (score processing device)

Claims

1. In a gaming information management system for an amusement center where multiple types of gaming machines are installed, a post collecting means for collecting player posts on a predetermined SNS site regarding a new gaming machine model during a predetermined period before and after the date on which the new gaming machine model is introduced; A polarity determination means for extracting words contained in the posts collected by the post collection means and determining a polarity indicating at least whether each word is positive or negative for each model; a polarity ratio calculation means for calculating a ratio of the polarities in all the posts as a polarity ratio for each model based on a result of the determination by the polarity determination means; an operation data calculation means for calculating operation data indicating an actual operation status after the new type of gaming machine is introduced; A gaming information management system characterized by comprising a learning model generation means that performs machine learning using the polarity ratio and operation data for each model as training data, and generates a learning model for predicting the operation status of the gaming machine.

2. the operation data calculation means calculates, as the operation data, the number of players who have played the gaming machine within a predetermined period after the introduction of the gaming machine; The gaming information management system described in claim 1, characterized in that the learning model generation means performs machine learning using the polarity ratio and the number of players as training data, and generates a learning model for predicting the number of players when a specific model is introduced.

3. the operating data calculation means calculates, each time a certain period of time has elapsed since the introduction of the gaming machine, an average gaming value consumed by one gaming machine during one business day during the certain period for each type of gaming machine, and calculates the period until the average falls below a reference value as an operating life period; The game information management system described in claim 1, characterized in that the learning model generation means performs machine learning using the polarity ratio and the operational life span as training data to generate a learning model for predicting the operational life span of a specific model.

4. The gaming information management system according to any one of claims 1 to 3, characterized in that, when a new model of gaming machine is newly introduced, a prediction means is provided for predicting the operating status of the new model of gaming machine after its introduction by inputting the polarity ratio for the new model into the learning model.

Images