CN104866699A - Intelligent online game data analysis method - Google Patents
Intelligent online game data analysis method Download PDFInfo
- Publication number
- CN104866699A CN104866699A CN201410063048.7A CN201410063048A CN104866699A CN 104866699 A CN104866699 A CN 104866699A CN 201410063048 A CN201410063048 A CN 201410063048A CN 104866699 A CN104866699 A CN 104866699A
- Authority
- CN
- China
- Prior art keywords
- behavior
- player
- difference
- game
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an intelligent online game data analysis method. The method comprises steps as follows: user data are collected, fluctuation conditions of behavior of players during game playing are established, a player behavior fluctuation curve is established, then player behavior characteristic calculation and difference detection are performed, characteristics of behavior of the players for two events as well as differences among different behavior are calculated, so that player behavior change and the correlation degree of the two events when the players play games can be monitored directly, when the player behavior changes, a system can give an alarm automatically and regulate the sequence of game pushing behavior and the sequence of events according to an MI curve and Di difference change, the r value is regulated automatically according to a player behavior difference curve, and the S value is regulated to be adapted to behavior requirements of the players. With the adoption of the method, the precision of background regulation and control cannot be affected, and the speed of the trend that user preference can change slowly as time goes on can be reflected.
Description
Technical field
The present invention relates to network data analysis field, particularly a kind of online game intelligent data analytical approach.
Background technology
At present, under the large data platform of network game user supports, system carries out the collection of information data, analysis, filtration, integration automatically, customizes easily, object that intelligence and the new generation network that individual differenceization is served are played to reach for a large amount of network game player simultaneously.
But the method for available data analysis has the following disadvantages, need to improve:
Along with the surge of number of users, the development in epoch, there is huge change in the demand of Internet user, how to be each customization game content accurately, and the game experiencing allowing each user obtain oneself the best is the direction of research always.The present invention establishes a game background intelligent adjustment and control system, can help the behavior of game developer's statistical study mass users, simultaneously according to user behavior Automatic adjusument, for each user provides game experiencing perfect as far as possible.
But this system yet also exists some shortcomings, first be in foundation, when quantizing user's Sample Storehouse, the method of the statistics adopted is mainly based on the game content of user, and the software application time difference of actual user's individuality is relatively large, game content is also subject to extraneous many factors of playing, therefore can some effects backstage regulation and control precision.
Next is that As time goes on the hobby of user may occur that slow change is (as work change, living environment changes, the game hobby change that marital status change etc. causes), the data model statistics of game is difficult to reflect " slowly changing " trend speed, therefore, although sorted out by feature extraction mode in sampling for a long time, but the result analyzed be merely able to be based on existing regulate and control, and be the reference of developer's subsequent development content providers tropism, entirely accurate can not go out dope user behavior trend.How can Accurate Prediction player behavior move towards, this is also ensuing primary study content.
In view of this, this area inventor, for the problems referred to above, have developed a kind of online game intelligent data analytical approach.
Summary of the invention
The invention provides a kind of online game intelligent data analytical approach, overcome the difficulty of prior art, the precision of backstage regulation and control can not be affected, and " slowly change " the trend speed that can reflect that As time goes on the hobby of user may go out.
Present invention employs following technical scheme:
The invention provides a kind of online game intelligent data analytical approach, comprise the following steps:
(1) the statistical sample storehouse behavior curve of cyclical fluctuations, when user enters game, system can collect all game behaviors such as user gradation, online hours, Offtime, preference task type, social relationships automatically, subsequently each individual behavior can by statistics be vector a: X=[, ], after carrying out a period of time statistics to the behavior of player, a Variant statistical is carried out in the behavior can accumulated all players:
Wherein, H is behavior quantity, and W is player's sample size, and f (x, y) is player's behavior every day vector; With this, calculate the mean difference numerical value of player's behavior on the two, in like manner individuality is added up separately, the fluctuation situation of player's behavior in the process of playing games can be calculated, set up player's behavior curve of cyclical fluctuations with this;
(2) extract player's behavioural characteristic and Difference test, use Mutual information entropy to calculate:
Wherein, now the unit of entropy is behavior number;
(3) combination entropy of two events is calculated:
(4) mutual information MI is calculated:
(5) behavior difference and MI are business, obtaining new difference is:
Wherein, N is the number of the time interval divided, and Hi (j) is the probability occurred in i-th behavior in jth day;
(6) the entropy HIi calculating i-th behavior is:
In like manner calculate the entropy of the i-th-1 behavior, then ask this two joint distribution rates, obtain the value of MIi, thus finally obtain difference Di; Wherein, Di reflects player for the otherness between the feature of behavior between two events and different behavior, thus can visual supervisory control player Behavioral change and player correlation degree when playing between two events, similarity degree; And
(7) when change appears in player's behavior, difference according to MI curve and Di changes, system understands automatic alarm and adjustment pushes the order of game behavior and the order of event, to the distribution of characteristic quantity Gaussian function Modling model, suppose to meet average the Gaussian distribution N (μ of to be μ variance be σ, σ), and to set threshold value be S=μ+r σ
Here, r is a parameter regulating flase drop number, and the establishing method of r is
Wherein, X is the MI average data of yesterday, and Y is the MI data occurred today, and T is the time, can according to player's behavior difference curve based on this system, and adjustment r value automatically, thus adjustment S value, adapt to the behavioral requirements of player.
Preferably, in described step (1), the fluctuation situation of player's behavior in the process of playing games, comprises the fluctuation situation of individual player or the fluctuation situation of all players.
Preferably, in described step (4), when the value of MI is larger, then the independence that Two Variables is mutual is less, and correlativity is larger.
Owing to employing above technology, online game intelligent data analytical approach of the present invention can not affect the precision of backstage regulation and control, and " slowly change " the trend speed that can reflect that As time goes on the hobby of user may go out.
The present invention is further illustrated below in conjunction with drawings and Examples.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of online game intelligent data analytical approach of the present invention;
Fig. 2 is the MI value change curve in the embodiment of the present invention;
Fig. 3 is in embodiments of the invention after use intelligent network game background system a period of time, the MI value curve of player; And
Fig. 4 is Precision-Recall distribution plan in embodiments of the invention.
Embodiment
Specific embodiments of the invention are introduced below by Fig. 1 to 4.
First embodiment
Fig. 1 is the process flow diagram of online game intelligent data analytical approach of the present invention.When method of the present invention being applied to intelligent game " rivers and lakes ", as shown in Figure 1, a kind of online game intelligent data analytical approach of the present invention, comprises the following steps:
The statistical sample storehouse behavior curve of cyclical fluctuations.When user enters intellectuality game " rivers and lakes " of giant's network research and development, system can collect all game behaviors such as user gradation, online hours, Offtime, preference task type, social relationships automatically.Subsequently each individual behavior can by statistics be vector a: X=[, ... ], after carrying out a period of time statistics to the behavior of player, a Variant statistical is carried out in the behavior can accumulated all players
Wherein H is behavior quantity, and W is player's sample size, and f (x, y) is player's behavior every day vector; The method calculates the mean difference numerical value of player's behavior on the two, in like manner adds up separately individuality, can calculate the fluctuation situation of player's behavior in the process of playing games
So far, set up player's behavior curve of cyclical fluctuations, for individual and entirety is all effective.
Extract player's behavioural characteristic and Difference test.Mutual information entropy is used to calculate.
Now the unit of entropy is behavior number.
Calculate the combination entropy of two events:
Calculate mutual information MI:
The value of MI is larger, and illustrate that the mutual independence of this Two Variables is less, so their correlativity is larger.
Behavior difference and MI are business, and obtaining new difference is:
Wherein N is the number of the time interval divided, and Hi (j) is the probability occurred in i-th behavior in jth day, is so again to calculate the entropy HIi of i-th behavior:
(formula 3.9)
In like manner calculate the entropy of the i-th-1 behavior, then ask this two joint distribution rates, obtain the value of MIi.Thus finally obtain difference Di.
Di fully reflects player for the otherness between the feature of behavior between two events and different behavior, thus can visual supervisory control player Behavioral change and player correlation degree when playing between two events, similarity degree.
Self-adapting data computing and intelligent learning.
When change appears in player's behavior, according to the difference change of MI curve and Di, system can automatic alarm and adjustment pushes the order of game behavior and the order of event, and namely oneself's adjustment is with the Behavioral change demand adapting to player.
To the distribution of characteristic quantity Gaussian function Modling model, suppose to meet average the Gaussian distribution N (μ, σ) of to be μ variance be σ, and set threshold value and be
S = μ + rσ
Here, r is a parameter regulating flase drop number.The establishing method of r is
Wherein X is the MI average data of yesterday, and Y is the MI data occurred today, and T is the time.Can according to player's behavior difference curve based on this system, adjustment r value automatically, thus adjustment S value, adapt to the behavioral requirements of player.
Data have been based upon on the large data processing basis of giant, can provide before using intellectualizing system and after intellectualizing system here, the player's behavior the change of divergence drawn:
Before use, as shown in Figure 2, wherein, transverse axis is the date to curve, and vertical pivot is MI value, and this data sampling comes from the journey whole district 500 days, and all of (in January, 2006 is in May, 2007) enliven player's behavioral data.Can see before use, the change of player's behavior every daily fluctuation is violent, player is described, and behavior every day difference is huge in gaming, and system interior function of can not carrying out playing regulates and controls in time, cause player after having played game 200 days, occur a large amount of player's wastage (change of MI value acutely), after a large amount of player is run off, residue player continues game, and MI value is smooth-out, but continues over time to occur big ups and downs.This curve conforms to the actual traffic-operating period of game.
As shown in Figure 3, after use intelligent network game background system a period of time, the MI value curve of player, wherein, transverse axis is the date, and vertical pivot is MI value.This data sampling is the data that journey will amount to 2000 days 2008 to 2013 end of the year, and data fluctuations amplitude on average declines about 20 ~ 30%, and upgrades on a large scale in game, game, during balance adjustment, all in curve, obtains embodiment.But after the adjustment, player's daily behavior is smooth-out, less appearance is extensive, and change for a long time, overall player gaming experience is significantly enhanced.
As shown in Figure 4, Precision-Recall distribution plan in embodiments of the invention, wherein, transverse axis is accuracy rate, and vertical pivot is recall rate.Visible investigate for different groups after, player is all better for auto-control repercussion.After years of researches experiment and improvement, this system is comprehensive in " rivers and lakes " game.
In summary, online game intelligent data analytical approach of the present invention can not affect the precision of backstage regulation and control, and " slowly change " the trend speed that can reflect that As time goes on the hobby of user may go out.
Above-described embodiment is only for illustration of technological thought and the feature of this patent, its object is to enable those skilled in the art understand the content of this patent and implement according to this, the scope of the claims of this patent only can not be limited with the present embodiment, namely the equal change done of all spirit disclosed according to this patent or modification, still drop in the scope of the claims of this patent.
Claims (3)
1. an online game intelligent data analytical approach, is characterized in that, comprises the following steps:
(1) the statistical sample storehouse behavior curve of cyclical fluctuations, when user enters game, system can collect all game behaviors such as user gradation, online hours, Offtime, preference task type, social relationships automatically, subsequently each individual behavior can by statistics be vector a: X=[, ], after a period of time statistics is carried out to the behavior of player, a Variant statistical is carried out to the behavior that all players accumulate:
Wherein, H is behavior quantity, and W is player's sample size, and f (x, y) is player's behavior every day vector; With this, calculate the mean difference numerical value of player's behavior on the two, in like manner individuality is added up separately, the fluctuation situation of player's behavior in the process of playing games can be calculated, set up player's behavior curve of cyclical fluctuations with this;
(2) extract player's behavioural characteristic and Difference test, use Mutual information entropy to calculate:
Wherein, now the unit of entropy is behavior number;
(3) combination entropy of two events is calculated:
(4) mutual information MI is calculated:
(5) behavior difference and MI are business, obtaining new difference is:
Wherein, N is the number of the time interval divided, and Hi (j) is the probability occurred in i-th behavior in jth day;
(6) the entropy HIi calculating i-th behavior is:
In like manner calculate the entropy of the i-th-1 behavior, then ask this two joint distribution rates, obtain the value of MIi, thus finally obtain difference Di; Wherein, Di reflects player for the otherness between the feature of behavior between two events and different behavior, thus can visual supervisory control player Behavioral change and player correlation degree when playing between two events, similarity degree; And
(7) when change appears in player's behavior, difference according to MI curve and Di changes, system understands automatic alarm and adjustment pushes the order of game behavior and the order of event, to the distribution of characteristic quantity Gaussian function Modling model, suppose to meet average the Gaussian distribution N (μ of to be μ variance be σ, σ), and to set threshold value be S=μ+r σ
Here, r is a parameter regulating flase drop number, and the establishing method of r is
Wherein, X is the MI average data of yesterday, and Y is the MI data occurred today, and T is the time, can according to player's behavior difference curve based on this system, and adjustment r value automatically, thus adjustment S value, adapt to the behavioral requirements of player.
2. online game intelligent data analytical approach as claimed in claim 1, is characterized in that: in described step (1) that the fluctuation situation of player's behavior in the process of playing games comprises the fluctuation situation of individual player or the fluctuation situation of all players.
3. online game intelligent data analytical approach as claimed in claim 1 or 2, it is characterized in that: in described step (4), when the value of MI is larger, then the independence that Two Variables is mutual is less, and correlativity is larger.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410063048.7A CN104866699B (en) | 2014-02-25 | 2014-02-25 | A kind of online game intelligent data analysis method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410063048.7A CN104866699B (en) | 2014-02-25 | 2014-02-25 | A kind of online game intelligent data analysis method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104866699A true CN104866699A (en) | 2015-08-26 |
| CN104866699B CN104866699B (en) | 2017-07-14 |
Family
ID=53912524
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410063048.7A Active CN104866699B (en) | 2014-02-25 | 2014-02-25 | A kind of online game intelligent data analysis method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104866699B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105536251A (en) * | 2015-12-15 | 2016-05-04 | 云南大学 | Automatic game task generation method based on user quality of experience fluctuation model |
| CN105956873A (en) * | 2016-04-08 | 2016-09-21 | 杭州碧游信息技术有限公司 | Game data intelligent analysis method |
| CN109464803A (en) * | 2018-11-05 | 2019-03-15 | 腾讯科技（深圳）有限公司 | Virtual objects controlled, model training method, device, storage medium and equipment |
| CN110678239A (en) * | 2017-10-10 | 2020-01-10 | 谷歌有限责任公司 | Distributed sample-based game profiling and game API platform supporting third party content with game metadata and metrics |
| US11654354B2 (en) | 2018-04-02 | 2023-05-23 | Google Llc | Resolution-based scaling of real-time interactive graphics |
| US11662051B2 (en) | 2018-11-16 | 2023-05-30 | Google Llc | Shadow tracking of real-time interactive simulations for complex system analysis |
| US11701587B2 (en) | 2018-03-22 | 2023-07-18 | Google Llc | Methods and systems for rendering and encoding content for online interactive gaming sessions |
| US11813521B2 (en) | 2018-04-10 | 2023-11-14 | Google Llc | Memory management in gaming rendering |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101753381A (en) * | 2009-12-25 | 2010-06-23 | 华中科技大学 | Method for detecting network attack behaviors |
| JP2010231524A (en) * | 2009-03-27 | 2010-10-14 | Renesas Electronics Corp | Data analysis method and data analysis program |
| CN102685766A (en) * | 2012-05-13 | 2012-09-19 | 西华大学 | Wireless network flow prediction method based on local minimax probability machine |
| CN103020976A (en) * | 2012-12-31 | 2013-04-03 | 中国科学院合肥物质科学研究院 | Method and system for registering three-dimensional medical images on basis of weighted fuzzy mutual information |
-
2014
- 2014-02-25 CN CN201410063048.7A patent/CN104866699B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010231524A (en) * | 2009-03-27 | 2010-10-14 | Renesas Electronics Corp | Data analysis method and data analysis program |
| CN101753381A (en) * | 2009-12-25 | 2010-06-23 | 华中科技大学 | Method for detecting network attack behaviors |
| CN102685766A (en) * | 2012-05-13 | 2012-09-19 | 西华大学 | Wireless network flow prediction method based on local minimax probability machine |
| CN103020976A (en) * | 2012-12-31 | 2013-04-03 | 中国科学院合肥物质科学研究院 | Method and system for registering three-dimensional medical images on basis of weighted fuzzy mutual information |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105536251A (en) * | 2015-12-15 | 2016-05-04 | 云南大学 | Automatic game task generation method based on user quality of experience fluctuation model |
| CN105956873A (en) * | 2016-04-08 | 2016-09-21 | 杭州碧游信息技术有限公司 | Game data intelligent analysis method |
| CN110678239A (en) * | 2017-10-10 | 2020-01-10 | 谷歌有限责任公司 | Distributed sample-based game profiling and game API platform supporting third party content with game metadata and metrics |
| US11684849B2 (en) | 2017-10-10 | 2023-06-27 | Google Llc | Distributed sample-based game profiling with game metadata and metrics and gaming API platform supporting third-party content |
| CN110678239B (en) * | 2017-10-10 | 2024-02-09 | 谷歌有限责任公司 | Distributed sample-based game profiling with game metadata and metrics and game API platform supporting third party content |
| US11701587B2 (en) | 2018-03-22 | 2023-07-18 | Google Llc | Methods and systems for rendering and encoding content for online interactive gaming sessions |
| US11654354B2 (en) | 2018-04-02 | 2023-05-23 | Google Llc | Resolution-based scaling of real-time interactive graphics |
| US11813521B2 (en) | 2018-04-10 | 2023-11-14 | Google Llc | Memory management in gaming rendering |
| CN109464803A (en) * | 2018-11-05 | 2019-03-15 | 腾讯科技（深圳）有限公司 | Virtual objects controlled, model training method, device, storage medium and equipment |
| CN109464803B (en) * | 2018-11-05 | 2022-03-04 | 腾讯科技（深圳）有限公司 | Virtual object control method, virtual object control device, model training device, storage medium and equipment |
| US11662051B2 (en) | 2018-11-16 | 2023-05-30 | Google Llc | Shadow tracking of real-time interactive simulations for complex system analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104866699B (en) | 2017-07-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104866699A (en) | Intelligent online game data analysis method | |
| Albert et al. | On the importance of intraspecific variability for the quantification of functional diversity | |
| CN108905209B (en) | Method and system for detecting plug-in game, electronic equipment and storage medium | |
| Sonkoly et al. | Both mass ratio effects and community diversity drive biomass production in a grassland experiment | |
| CN109472004A (en) | Comprehensive estimation method, the apparatus and system that climate change and mankind's activity influence Hydrologic Drought | |
| CN103731738A (en) | Video recommendation method and device based on user group behavioral analysis | |
| CN104102875A (en) | Software service quality monitoring method and system based on weighted naive Bayes classifier | |
| Mediero et al. | Probabilistic calibration of a distributed hydrological model for flood forecasting | |
| CN111510783B (en) | Method, device, electronic equipment and storage medium for determining video exposure | |
| CN105184400A (en) | Tobacco field soil moisture prediction method | |
| CN103605878A (en) | General blood glucose prediction method based on data modeling and model transplanting | |
| CN109508429A (en) | Personalized adaptive learning recommended method based on teaching platform big data analysis | |
| Khan et al. | How do tree competition and stand dynamics lead to spatial patterns in monospecific mangroves? | |
| Marquis et al. | Variability in frost occurrence under climate change and consequent risk of damage to trees of western Quebec, Canada | |
| CN109635192A (en) | Magnanimity information temperature seniority among brothers and sisters update method and platform towards micro services | |
| Fahlvik et al. | Simulating the influence of initial stand structure on the development of young mixtures of Norway spruce and birch | |
| CN110738565A (en) | Real estate finance artificial intelligence composite wind control model based on data set | |
| Pretzsch et al. | Coarse root–shoot allometry of Pinus radiata modified by site conditions in the Western Cape province of South Africa | |
| CN108985493A (en) | A kind of ground class variation prediction method based on self-adapting changeable filter | |
| CN112330368B (en) | Data processing method, system, storage medium and terminal equipment | |
| CN103823843A (en) | Gauss mixture model tree and incremental clustering method thereof | |
| Wang et al. | player2vec: A Language Modeling Approach to Understand Player Behavior in Games | |
| Sudin et al. | Cycling performance prediction based on cadence analysis by using multiple regression | |
| CN106846064B (en) | Software potential ordering method based on co-occurrence relation | |
| Garcia | Reverse causality in size-dependent growth |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |