KR20180056151A - Self learnig system and method for electronic card game using the graph database approach - Google Patents

Abstract

The present invention relates to a self learning system and a method of an electronic card game using a graph database manner. The present invention provides an algorithm for providing an electronic card game, and does not directly create and reflect an interaction relationship between variables of each electronic card object. In the present invention, an electronic card game system compares and judges the variables of the electronic card object itself under a minimum game rule, and stores and learns all the game progress history and improves the algorithm by itself. Accordingly, the present invention relates to the self-learning system and the method of the electronic card game using a graph database manner which can determine the game progress direction on the basis of the learned game progress history at the later game progress.

Description

[0001] SELF LEARNING SYSTEM AND METHOD FOR ELECTRONIC CARD GAME USING THE GRAPH DATABASE APPROACH [0002]

The present invention relates to a self-learning system and method of an electronic card game using a graph database method, and provides an algorithm for providing an electronic card game, in which an interaction relationship between variables of each electronic card object is directly The electronic card game system compares and judges the variables of the electronic card object by itself and saves and learns all the game progress histories and refines the algorithm by itself, And a self-learning system and method of an electronic card game using a graph database method capable of determining a game progress direction based on a progress history.

Generally, as the game's artificial intelligence algorithm is pre-programmed by the developer and registered in the database, limited artificial intelligence play is possible only by a pre-registered algorithm. In order to solve or improve the error of such algorithm, It becomes cumbersome to carry out direct programming.

However, since the amount of game content is enormous and participants have a large number of users, variables to be considered for generating the algorithm of the corresponding game are also enormous, Time, development cost, and the like, as well as increasing the complexity of the game, thereby increasing the error occurrence rate in the game.

In particular, in-game data can be changed in real time by an algorithm in a complicated game, and it is difficult for a person to optimize correspondingly all of these problems.

Accordingly, the present inventor has provided an algorithm for providing an electronic card game in order to solve various problems and limitations of the above-mentioned conventional game artificial intelligence realization method, so that an interaction relation between variables of each electronic card object The electronic card game system compares and judges the variables of the electronic card object by itself and stores and learns all game progress histories under the minimum game rules, thereby improving the algorithm itself, Learning system and method of the electronic card game using the graph database method capable of determining the direction of the game on its own based on the learned game progress history.

Korean Patent No. 10-1231798

It is an object of the present invention to provide an algorithm for providing an electronic card game, and it is an object of the present invention to provide an algorithm for providing an electronic card game system, in which an interaction relationship between variables of each electronic card object is directly created and reflected, A graph database capable of determining a game progress direction by itself based on the game progress history learned at the time of game progress by comparing and determining the variables of the electronic card object by itself and storing and learning all the game progress histories, And to provide a self-learning system and method for an electronic card game using the method.

A self-learning method of an electronic card game using a graph database method according to an embodiment of the present invention includes storing a game progress history in a database for a game progress history, displaying one or more first card objects presented in a game field, Inquiring the past game progress history between the second card objects in the game progress history database and determining the second card object to be presented next time based on the odds set by the administrator.

In one embodiment, the step of storing the game progress history includes storing at least one of an identification code of at least one card object presented in a game field, an attack power value, a defense power value, a fitness value, a cost value, And may store at least one of a numerical value, an attack history, a defense history, a buff history, a debuff history, a victory number, and a defeat number.

In one embodiment, the step of inquiring includes the steps of: attacking, defending, buffing, debuffing, winning, defeating, receiving damages, and receiving damage of the second card object for each of the one or more first card objects presented And inquiring at least one of the provided damage amount values.

In one embodiment, the winning percentage may be set based on a card game use time of a user who plays a card game, a frequency of using the card game, and a frequency of using the charged item in the card game.

In one embodiment, the step of determining the second card object to be presented next may include the step of determining that the second card object performs an attacking action, a defensive action, a buffing action, or a debuff action for each of the one or more first card objects presented Calculating a damage amount value that has been received, and determining any one of the second card object to be presented next time and the one or more first card objects presented according to the set winning percentage.

The self-learning system of the electronic card game using the graph database system according to another embodiment of the present invention includes a database for a game progress history storing a game progress history, at least one first card presented in the game field in the game progress history database And a determination unit for determining a second card object to be presented next time based on a winning ratio set by the manager and the inquiry unit for inquiring the past game progress history between the object and the second card object to be presented next time.

In one embodiment, the database for the game progress history includes at least one of an identification code, a damage value, a defense value, a fitness value, a cost value, a special ability value, a damage value, At least one of an attack history, a defense history, a buff history, a debuff history, a victory number and a defeat number.

In one embodiment, the inquiry unit is configured to calculate the attack history, the defense history, the buff history, the debuff history, the victory number, the defeat number, the damage amount value and the received damage amount of the second card object for each of the one or more first card objects presented At least one of the damage amount values can be inquired.

In one embodiment, the determining unit includes a calculating unit for calculating a damage amount value that is received when the second card object performs an attacking action, a defending action, a buffing action, or a debuffing action for each of the one or more first card objects presented in advance And the calculating unit may determine either the second card object to be presented next time or the first or more card object presented in advance according to the set winning rate.

According to one aspect of the present invention, in providing an algorithm for providing an electronic card game, it is possible to provide an algorithm for providing an electronic card game under a minimum game rule, rather than directly creating and reflecting the interaction relationship between the variables of each electronic card object, The game system compares and judges the variables of the electronic card object by itself and stores and learns all the game progress histories and improves the algorithm by itself so that the game progress direction itself can be determined based on the learned game progress history in the future game progress, Even if the contents in the game are further expanded and the number of users is increased, there is an advantage that the algorithm can be optimized and improved by itself under the basic game rule without directly intervening the developer.

FIG. 1 is a diagram illustrating a configuration of a self-learning system 100 of an electronic card game using a graph database system according to an embodiment of the present invention.
FIG. 2 is a view showing a correlation between card objects according to the present invention.
3 is a diagram illustrating a series of processes of self-learning an electronic card game through a self-learning system 100 of an electronic card game using a graph database system according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in order to facilitate understanding of the present invention. However, the following examples are provided only for the purpose of easier understanding of the present invention, and the present invention is not limited by the examples.

FIG. 1 is a diagram illustrating a configuration of a self-learning system 100 for an electronic card game using a graph database system according to an embodiment of the present invention. FIG.

1 and 2, the self-learning system 100 of an electronic card game using the graph database system according to the present invention mainly includes a database 110 for a game progress history, a query unit 120, and a determination unit 130 And the like.

First, the game progress history database 110 stores all game progress histories related to the progress of the electronic card game.

Here, the game progress history refers to, for example, an identification code of one or more card objects presented in a game field, an attack power value, a defense power value, a fitness value, a cost value, a special ability value, , The attack history, the defense history, the BUFF history, the DEBUFF history, the victory frequency, and the defeat frequency.

Meanwhile, the game progress history database 110 may store all the logs related to the card object behaviors, which are generally performed in the online electronic card game, in addition to the types of the game progress history described above. .

Next, the inquiry unit 120 acquires past game progress history between the at least one first card object 1 presented in the game field in the game progress history database 110 and the second card object 2 presented next time It can perform the role of inquiry.

More specifically, the inquiry unit 120 acquires the attack history, the defense history, the buff history, the debuff history, the number of victories, the number of defeats, the number of victories, The game progress database 110 may inquire at least one of the received damage value and the provided damage value.

Here, the first card object 1 and the second card object 2 can give meaning to the electronic card itself as a card object in an electronic card game. The card object includes a card name, an attack power of the card, The card's defense, the cost of the card, and the card's special abilities. In other words, such a card object can be referred to as node data.

In addition, each card object may perform a variety of interrelated actions on other card objects such as an attack, defensive, buff, debuff, or other special action. In addition, the number of win / lose times of the game for each card object can be included. In other words, the mutually related actions can be referred to as edge data.

Accordingly, as the electronic card game progresses, not only the interaction of each card object (by node data) but also the data of each of the correlated actions (by edge data) is continuously recorded in the game progress database 110 ). ≪ / RTI > Based on the accumulated data, the self-learning system 100 of the electronic card game using the graph database method according to the present invention can continuously optimize its own algorithm.

In particular, in the case of the inquiry unit 120, since all the past past game progress histories are stored, the card object to be presented next to the card object presented in the game field by the user at the time of the electronic card game with the user If it is determined that the game progress history has been stored in the game progress database 110, the inquiry unit 120 may store the accumulated game progress history between the card object present in the current game field and the card object to be presented next, among the vast amount of past game progress history previously stored in the game progress database 110 It is possible to select a card object to be presented next based on the inquiry.

Next, the determination unit 130 determines the second card object to be presented next time, based on the winning rate set by the administrator.

More specifically, the determination unit 130 determines whether or not the damage amount value (hereinafter referred to as " damage amount ") that the second card object 2 has received as a result of performing an attacking, defending, And a second card object 2 to be presented next time in accordance with the odds set by the administrator, and one or more first card objects 1 that are presented in advance have.

For example, to calculate the damage value received here is as follows.

When the fitness value of the second card object 2 is reduced by the attack value of the first card object 1 as the attacking action is applied to the second card object 2 in the first card object 1, , It can be determined that the amount of damage to the second card object 2 has been generated based on this, and the value of the amount of damage to the second card object 2 can be the received damage value.

In the case of the electronic card game, the fitness value of each card object can be changed in real time according to the received damage value because the game can be terminated by defeating the player when the fitness value of the user is zero.

Accordingly, since the fitness value of each card object can be changed in real time by mutual attack, defensive action, buff act or debuff act for each card object, the calculation unit 131 can calculate the damage value It is possible to judge whether or not the electronic card game has won or lost based on the change amount of the physical strength value.

The winning rate set by the administrator may mean a condition that is set based on the card game use time of the user who plays the card game, the frequency of using the card game, and the frequency of using the charged item in the card game. (For example, 60%) set for each user who uses the card game after storing the win / loss ratio of each game according to the object-based interaction.

This will be described with reference to an example of the following game progress.

In one embodiment, four first card objects corresponding to A, B, C, and D are presented in the game field of the counterpart (user), and in the artificial intelligence card game system confronting them, Suppose a card object is presented.

In case of the E card object possessed by the AI card game system, the attack power value is 5, the physical strength value is 1, the A card object has the attack power value of 2, the physical strength value is 7, 1, the physical strength value is 1, the attack power value is 1, the physical strength value is 5 for the C card object, the attack power value is 2, and the physical strength value is 7 for the D card object.

If an A card object is attacked using an E card object, the attack power value of the A card object is 2, the health value is 2, and the E card object can be destroyed. However, if you are attacked from A, B, C, D card objects in the future, you will receive the damage value corresponding to the total 9 attack value.

If the B card object is attacked using the E card object, the B card object and the E card object may be destroyed. However, if attacked from A, C, and D card objects in the future, the damage value will be equal to 8 damage points.

If the C card object is attacked using the E card object, the C card object and the E card object may be destroyed. However, if you are attacked from A, B, and D card objects in the future, you will receive the damage value corresponding to the total 5 damage figures.

If an E card object is used to attack a D card object, only the E card object can be destroyed. If the card object is attacked by A, B, C, and D card objects, assuming that the card object has a special ability that increases the next attack value by 3 when attacked, A numerical value is generated.

That is, when the second card object 2 is presented to the various first card objects 1 presented in a line and interacted with the specific first card object 1, different damage values are generated in the next game .

Therefore, the determination unit 130 determines whether or not the second card object 2 to be presented, as well as the second card object 2 to be presented, according to the first to third conditions set by the administrator among the different damage value values 9, 8, 5, It becomes possible to select with the first card object 1 to act in correlation with the card object 2. [

For example, when the condition is set by the manager to win the 100% victory probability of the AI game progress system, the second card object 2 and the second game object 2 are set to be the smallest damage value among the different damage values, It becomes possible to select the first card object 1 that will interact with the card object 2.

For example, when the condition is set such that the AI game progress system is won by a manager with a winning probability of 30%, the second card object 2 and the second card object 2 are set to be the largest damage value among the different damage values, 2 < / RTI > card object < RTI ID = 0.0 > 1 < / RTI >

This embodiment is merely an example for explaining the mutually related action between the first card object 1 and the second card object 2 and is not limited to the attack value, the physical strength value, the received damage value, And the like are not limited.

As a result, if the specified physical strength value of the opponent is set to 0, the odds of the game is increased. If the opponent's damage value is larger, the odds of the opponent are lowered. And all of these processes are stored and accumulated in the database 110 for the game progress history, so that they will be used as reference data for future electronic card games.

Meanwhile, in one embodiment, since there are many game progress histories stored and accumulated in the game progress history database 110, the game progress history database 110 stores and stores the game progress history between the user and the user. And the odds data and the algorithm according to the interrelated action of each card object may be accumulated starting from the game progress history between the user and the user.

Next, a series of processes of self-learning of the electronic card game through the self-learning system 100 of the electronic card game using the graph database method will be sequentially examined through FIG. 3.

3 is a diagram illustrating a series of processes of self-learning an electronic card game through a self-learning system 100 of an electronic card game using a graph database system according to an embodiment of the present invention.

First, all game progress histories are stored in the game progress history database 110 starting from the game progress history between the user and the user (S301). At this time, the one or more card object identification codes, the attack value, the defense value, the health value, the cost value, the special ability value, the damage value received, the damage value provided, the attack history, the defense history, The history, the number of victories, and the number of defeats.

Next, the past game progress history between the at least one first card object presented in the game field and the second card object to be presented next is inquired at the game progress history database 110 at step S302. At this time, At least any one of the attack history, the defense history, the buff history, the debuff history, the victory number, the defeat number, the damage amount value, and the damage amount value of the second card object for each of the first card objects is searched.

Next, the damage amount value that is received when the second card object performs the attacking action, the defensive action, the buffing action, or the debuffing action for each of the one or more first card objects presented is calculated (S303) According to any one of the preset first to third conditions, the first card object to be correlated with the second card object to be presented next and the second card object is determined (S304).

As described above, the self-learning method of the electronic card game using the graph database system according to the present invention is not to directly create and reflect the interaction relationship between the variables of each electronic card object, Since the card game system compares the variables of the electronic card object by itself and stores and learns all of the game progress histories and can expand the algorithm itself, it is possible to use the cards to be submitted next time based on the learned game progress history The object can be selected. In particular, since the odds can be adjusted so that the user can experience the optimal fun based on the odds set by the administrator, it is possible to optimize the algorithm suitable for each situation in real time.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the present invention as defined by the following claims It can be understood that

1: first card object
2: second card object
100: Self-learning system of electronic card game
110: Database for game progress history
120:
130:
131:

Claims (10)

Storing a game progress history in a database for the game progress history;
Querying a past game progress history between at least one first card object presented in a game field and a second card object to be presented next, in the game progress history database; And
And determining a second card object to be presented next time, based on a winning rate set by an administrator, the self-learning method of the electronic card game using the graph database method.
The method according to claim 1,
Wherein the step of storing the game progress history comprises:
The number of damage items received, the attack history, the defense history, the buff history, the number of the damage, Wherein at least one of a debuff history, a winning number, and a defeat number is stored in the graph database.
3. The method of claim 2,
Wherein the querying comprises:
At least one of an attack history, a defensive history, a buff history, a debuff history, a victory number, a defeat number, a damage amount value, and a damage amount value of the second card object for each of the at least one first card object presented The method comprising the steps of: (a) displaying a graph database of the electronic card game;
3. The method of claim 2,
The above-
Wherein said game program is set based on a card game use time of a user who plays a card game, a frequency of using a card game, and a frequency of using a charged item in a card game.
5. The method of claim 4,
Wherein the step of determining the second card object to be presented next includes:
Calculating a damage amount value that is received when the second card object performs an attacking action, a defense action, a buffing action, or a debuffing action for each of the one or more first card objects presented; And
And determining one of the second card object to be presented next time and the one or more first card objects presented in advance according to the set odds ratio. The electronic card game according to claim 1, Self - learning method.
A game progress history database storing the game progress history;
An inquiry unit for inquiring a past game progress history between at least one first card object presented in the game field in the game progress history database and a second card object to be presented next time; And
And a determination unit for determining a second card object to be presented next time based on a winning rate set by an administrator.
The method according to claim 6,
The game progress history database includes:
The number of damage items received, the attack history, the defense history, the buff history, the number of the damage, Wherein at least one of a debuff history, a winning number, and a defeat number is stored.
8. The method of claim 7,
Wherein,
At least one of an attack history, a defensive history, a buff history, a debuff history, a victory number, a defeat number, a damage amount value, and a damage amount value of the second card object for each of the at least one first card object presented Wherein the electronic card game is a self-learning system of electronic card game using a graph database system.
9. The method of claim 8,
The above-
Wherein the self-learning system is set based on a card game use time of a user who plays a card game, a frequency of using the card game, and a frequency of using a charged item in a card game.
10. The method of claim 9,
Wherein,
And a calculating unit for calculating a damage amount value that is received when the second card object performs an attacking action, a defending action, a buffing action, or a debuffing action for each of the one or more first card objects presented,
Characterized in that the calculating unit determines either the second card object to be presented next time or the one or more first card objects presented in advance according to the set winning percentage, system.

Images