CN100476819C - Data mining system based on Web and control method thereof - Google Patents
Data mining system based on Web and control method thereof Download PDFInfo
- Publication number
- CN100476819C CN100476819C CNB2006100226681A CN200610022668A CN100476819C CN 100476819 C CN100476819 C CN 100476819C CN B2006100226681 A CNB2006100226681 A CN B2006100226681A CN 200610022668 A CN200610022668 A CN 200610022668A CN 100476819 C CN100476819 C CN 100476819C
- Authority
- CN
- China
- Prior art keywords
- data
- mining
- web
- module
- user
- 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.)
- Expired - Fee Related
Links
Images
Abstract
The invention opens a data mining system based on Web, which mainly includes EJB server, Web server and database etc. The EJB server provides the interface between Web server and EJB layer, and implemented various data mining algorithms for different data and data mining tasks. The Web layer provides a user interactive interface to receive user inputs and display results of data mining and analysis. The system consists of several modules: authentication module, initialization module, data connection module, data visualization module, data pre-processing module, mining module, mining data indication module. The system provides on-line internet based data mining and result analysis services.
Description
Technical field
The present invention relates to the data mining technology field, be specifically related to a kind of data digging system and control method thereof based on Web.
Background technology
Data mining in brief, is exactly to be exactly from a large amount of incomplete real application data from data mining, extract lie in wherein, people are ignorant in advance but the information that comes in handy and the process of knowledge.Along with the development of computer hardware technology and the exploitation of various database, more data is collected in the computing machine with unprecedented speed, and its quantity and complexity are considerably beyond people's analysis ability.Therefrom find potential rule and information owing to lack effective instrument again, the mankind just have been absorbed in " rich data " and " knowledge of poorness " and the condition of the embarrassment of depositing.Like this, some important decision-makings often are not based on the mass data of collection, and are based on decision maker's intuition.Therefore, people wish to calculate function and help us to analyze data, understand data, therefrom find important data pattern or knowledge, help us to make a policy in fields such as commercial decision-making, science and medical researches, prediction development in future trend is so caused the generation of data mining technology.
Data mining has caused the extensive concern of domestic and international academia and industrial community as an infotech of extracting knowledge from mass data, and becomes a research focus in the computer realm.Simultaneously, it makes the software developer constantly develop and develops new Data Mining Tools in some successful Application aspect commercial.Because domestic enterprise can't specified data excavate the input risk of project and the income of expection, it is very slow that this makes domestic data mining use progress.Present data mining product mostly is to be provided by external large enterprises, as IBM; Buy and use a large amount of material resources of these product needed user efforts and manpower; These softwares are generally towards Fat Client (C/S) design, and limited system resource makes that operating mass data on this basis when small business or company can reduce digging efficiency unavoidably, brings a series of inconvenience to client.And present data mining software all also is in the exploratory stage, also is not very ripe product.
Comparatively famous data digging system is the Weka system of Waikato university of New Zealand university exploitation at present, it is a Data Mining Tools bag that function is stronger, provide a cover complete data mining process: comprise that data connect, unified data objects processing, data pre-service, mining algorithm commonly used, excavate result's expression etc., its uses graphical interfaces and user interactions based on the GUI of Swing.But it also has certain defective and needs perfect place.The deficiency of Weka shows: 1, it needs user installation software and related hardware, uses inconvenient; 2, only be a learning prototype system, support the mining analysis of big data quantity is very difficult; 3, Weka shows the excavation result in the mode of text, causes non-senior professional to be difficult to understand.
Summary of the invention
Technical matters to be solved by this invention is how a kind of data digging system and control method based on Web is provided, it is online that this system can provide, data mining and interpretation of result service based on Internet, can make the user not need to drop under the situation of substantial contribution, obtain high-quality data mining service, and utilize implicit and its management of valuable information guiding excavate, thereby for the business decision of company and enterprise provides more reasonable and Useful Information, and user side need not be installed any software and hardware, just can directly carry out data mining, and provide the data mining process of easy understanding and result's visualized graphs to show.
First technical matters proposed by the invention is to solve like this: construct a kind of data digging system based on Web, it mainly comprises assemblies such as EJB (Enterprise JavaBean) server, Web server and database, wherein the EJB server provides the interface between Web service end and the EJB layer, and realized the processing of data mining algorithm, the Web layer provides the interface of and user interactions, accept the user input, the mining process interactive interface is provided and shows the mining analysis result, it is characterized in that comprising following module:
Authentication module: is that the user carries out authentication by INTERNET to browser;
Initialization module:, the user of different stage is directed to the different pages to system initialization;
Data link block: the connection of various different data format data sources is provided, generates the database that mining algorithm can directly use, the interface of data object visit is provided;
Data visualization represents module: use the understandable media performance complex data and the correlationship thereof that can produce visual impression;
Data preprocessing module: provide source data is carried out pre-service, comprise the various processing modes of data scrubbing, integrated, conversion and reduction;
Mining model evaluation module: utilize different mining models that test data is carried out mining analysis, the Different Results that obtains is assessed, select corresponding mining model to carry out data mining according to assessment result;
Excavate processing module: the data mining algorithm processing mode is provided, carries out the data mining task that the user submits to;
Excavate display module as a result: the result of data mining is showed the user by media, provide intuitively, patterned excavation and analysis result.
According to the data digging system based on Web provided by the present invention, it is characterized in that EJB service end and Web service end are linked up by the Facade pattern.
According to the data digging system based on Web provided by the present invention, it is characterized in that, the Web service end be responsible for responding browser request, provide service for it, interface with Enterprise Java Bean container is provided, be used to represent html page and the importation of accepting the user, comprise JSP, Servlet and JavaBean assembly to client.
According to the data digging system based on Web provided by the present invention, it is characterized in that described data visualization represents module and excavates media in the display module as a result and can be in point and line chart, histogram, pie chart, network diagramming, Interactive Visualization, dynamic similation, the computer animation one or several.
According to the data digging system based on Web provided by the present invention, it is characterized in that described data preprocessing module and excavation processing module are provided with the interface that can increase new data algorithm at any time.
According to the data digging system based on Web provided by the present invention, it is characterized in that, database can be placed on the server of diverse location among the Internet, drives being connected of WEB end assembly and database by JDBC, supports the data file of client to upload to WEB and holds assembly; WEB end assembly obtains Data Post and becomes corresponding data object, and this data object is passed to the processing of applied business logic module; WEB end assembly can be placed on respectively on the different servers with the applied business logic module, also can be on same server, and both realize remote object invocation by RMI-IIOP mechanism; The data link block adopts multi-thread mechanism, and keeps synchronously.
According to the data digging system based on Web provided by the present invention, it is characterized in that system follows Struts1.1 framework bag and EJB2.0 standard based at the J2EE of Java 2 platform, adopt MVC three-layer architecture model; Business Logic mainly comprises EJB object, JavaBean; Key-course mainly comprises the Action action class based on Servlet; The view layer mainly is made of the JSP page; Application server and WEB server can adopt the J2EE application server of various standards; Wherein WEB end assembly can run on various computer systems; The application end assembly can run on various computer systems.
A kind of control method of the data digging system based on Web is characterized in that, comprises following steps:
(1), connect long-range or local data source, obtain corresponding data object by data-interface;
(2), show raw data by the data visualization module to remote user end;
(3), utilize data preprocessing module that raw data is handled;
(4), select suitable mining algorithm according to mutually deserved mining task;
(5), respective algorithms is carried out the parameter adjustment setting and reach suitable algorithm adjustment;
(6), utilize training data that corresponding algorithm model is trained, obtain the mining analysis model;
(7), utilize test data that the mining analysis model is assessed, whether select the new algorithm model of needs according to assessment result, perhaps the parameter of original algorithm model is done corresponding adjustment;
(8), utilizing the best mining model of assessment result in the above-mentioned steps that corresponding data object is carried out mining analysis handles;
(9), will excavate the result utilizes the medium and holds displaying to long-range or local user.
Control method according to the data digging system based on Web provided by the present invention, it is characterized in that, the data mining analysis service that native system can facilitate to all Internet users makes the user can obtain high-quality data mining, Analysis Service with minimum cost.
Control method according to the data digging system based on Web provided by the present invention is characterized in that, all of native system and user mutual and represent content all undertaken by browser.
Control method according to the data digging system based on Web provided by the present invention is characterized in that, the step that system responses user's page request is handled is as follows:
(1), controller Servlet receives the Http Request request from client, and is converted to the Event incident;
(2), the JavaBean that calls in the Model layer according to corresponding Event incident of controller Servlet begins to carry out business logic processing;
(3), the JavaBean in the Model layer realize in the middle of scheduling, the EJB assembly that calls in the Business Logic is realized service logic, the EJB assembly can be realized access and computing to business datum by database and algorithm controls engine;
(4), controller Servlet is according to service processing result, resolution path URL calls the corresponding JSP page;
(5), the correlation method in the JSP page invocation Business Logic is obtained data;
(6), JSP is according to data, generates html page, returns browser, carries out directly perceived, understandable mining analysis result's visual presentation.
Data digging system based on Web provided by the present invention can be provided in line, based on data mining and the interpretation of result service of Internet, can make the user not need to drop under the situation of substantial contribution like this, obtain high-quality data mining service, and utilize implicit and its management of valuable information guiding excavate, thereby for the business decision of company and enterprise provides more reasonable and Useful Information, guide its faster, better development, in intense market competition, occupy first chance.
Data digging system based on Web provided by the present invention, the user is hardly with any extra hardware and software investment, just can excavate, and from excavate the result, obtain to be directly used in business decision, but be hidden in data useful information behind at the enterprising line data of this system.The data mining service of this mode, popularization data mining that will be positive is in the application of commercial field and promote the development that it is good.It can carry out collective analysis to the various data of distribution isomery and handle, provide friendly data mining results to show, can also provide convenience and intelligentized interactive function to the user, be convenient to the more effective grasp mining process of user, fully understand the result with analysis mining.The more important thing is that can offer user-friendlyly, easily data mining process of understanding and result's visualized graphs is showed.Therefore, it is different from the excavation result's of traditional data mining product text representation mode, makes the user be easy to just can understand the result of data mining, reduces the misunderstanding and the omission of the effective information that produces owing to too much intermediate link.
Data digging system provided by the present invention is to provide the data mining service for the user on the internet: the various service modules (as association analysis, cluster analysis, classification analysis etc.) that the user can using system provides after by registration carry out data mining analysis to the raw data of oneself, a large amount of visualization functions is provided, make data analysis, pre-service and mining process have more human-computer interaction functions, and offer the multiple data visualization intuitively of user display result, allow the user be more readily understood and analysis mining information.Simultaneously, for the protection of user's data safety and data-privacy is considered, the user can utilize the native system analysis that the data that provide oneself are provided, can make sensitivity and significant data is in having in one's pocket of oneself, helps realizing the protection of private data.
The online service function of data digging system provided by the present invention is very powerful, has almost contained all main method that current data is excavated; System can handle several data source (various databases, several data file etc.) towards distributed development; System has function of keeping secret preferably for the raw data that the user submits to; And our the abundant visualization display module of exploitation allows the user can get more information about the result of excavation.
Data digging system provided by the present invention is to develop and design on the J2EE of Java 2 platform, and based on Struts1.1 framework bag and EJB2.0 standard, adopts MVC three-layer architecture model.Wherein Business Logic mainly comprises EJB object, JavaBean; Key-course mainly comprises the Action action class based on Servlet; The view layer mainly is made of the JSP page.Native system is developed based on distributed system, and application server and WEB server can adopt various standard J2EE application servers, can handle the data in the various Sybases.Wherein WEB end assembly can run on various computer systems; The application end assembly also can run on various computer systems.。
Our performance history is mainly carried out according to the prototyping mode of soft project, makes system constantly perfect by the mode of adding assembly.With the thin client's structure of present popular B/S is framework, and adopts OO distributed component development scheme, thereby guarantees high efficiency, security and the portability of this system's operation.
It is as follows based on the characteristics of the data digging system of Web that the present invention carried:
1, system, is separated service logic and client according to the B/S mode development based on the MVC three coating systems models of J2EE, thereby has alleviated the work load of client process machine greatly.Therefore, the decision maker of enterprise or company can analyze and formulate business strategy fast and efficiently in this system.
2, the weka data mining algorithm bag content of the system integration is very abundant, has almost contained the mining algorithm of present all main flows.The user can select corresponding algorithm to carry out data mining according to actual conditions.If run into news, because the design of system is based on prototype model development idea, we can add in the total system then at special sector or the new mining algorithm assembly of unit exploitation one cover fully, still are that function all is easy to control and expand from cost like this.
3, the visualized graphs exposition of Wa Jueing is based on powerful Java2D and Java3D technology, and with the popular up till now graph making project JfreeChart that increases income.This has guaranteed that system will provide powerful and visual and understandable visual excavation result to the user.The exploitation of data mining visualization component has huge promotion prospect, and we are encapsulated into visualization component among the different JavaBean when design, independently finish difference in functionality separately, and this has also demonstrated fully OO development idea.
4, the function of system realizes based on distributed, multithreading thought.At first the source of data distributes, and database can be placed on the server of diverse location among the Internet, drives being connected of WEB end assembly and database by JDBC, supports the data file of client to upload to WEB and holds assembly; WEB end assembly obtains Data Post and becomes corresponding data object, and this data object is passed to the processing of applied business logic module; The WEB end assembly of its subsystem can be placed on respectively on the different servers with the applied business logic module, also can be positioned at same application server, and both realize remote object invocation by RMI-IIOP mechanism.In addition, the transmission course of user's upload file adopts multi-thread mechanism, and keeps synchronously.This data upload process that has guaranteed different user does not have conflict fast.Application server itself also has the multiple line distance management function, and the user needn't the conflict of worry system in the process of handling data mining separately.
5, system is based upon on the security platform mechanism of Java2, on the one hand, can in time provide reasonably emergent solution to the fault of emerged in operation and pathological system; On the other hand, system adopts identity to land verification technique, and the data that the user uploads are carried out respective handling, thereby plays the purpose of protection user data and privacy.
Along with Internet the popularizing gradually of China, present client-based C/S software and stand-alone software forward change based on the software of B/S framework, we can say that following software development direction is a network-oriented, towards thin client's pattern.Native system is that the user provides complete data mining service with the online mode; Its operation platform is based upon perfect in shape and function and on the powerful J2EE, this has fully guaranteed the rationality and the integrality of system development; The network application of Java2 is very extensive, and the user can be smooth and be visited native system efficiently; Simultaneously the security mechanism of Java2 has guaranteed the security of system's operation and has broken down and rationally emergent solution that the boundary is normal, simultaneously the in addition appropriate secrecy provision of the native system data of the user being uploaded and storing by the mode of digital signature and certificate.
Description of drawings
Fig. 1 is the system assumption diagram of the data digging system based on Web provided by the present invention;
Fig. 2 is the workflow diagram of the data digging system based on Web provided by the present invention;
Fig. 3 is the module map of the data digging system based on Web provided by the present invention;
Fig. 4 is the interaction figure based on EJB and Web in the data digging system of Web provided by the present invention;
Fig. 5 is in the data digging system based on Web provided by the present invention
Pattern diagram.
Fig. 6 is the IPO figure of the data digging system based on Web provided by the present invention;
Fig. 7 is the operational flowchart of the data digging system based on Web provided by the present invention;
Fig. 8 is the logic diagram of the data digging system based on Web provided by the present invention;
Fig. 9 is provided by the present invention based on authentication module figure in the data digging system of Web;
Figure 10 is provided by the present invention based on system initialization module figure in the data digging system of Web;
Figure 11 is provided by the present invention based on data source link block figure in the data digging system of Web;
Figure 12 provided by the present inventionly represents module map based on data visualization in the data digging system of Web;
Figure 13 is provided by the present invention based on data preprocessing module figure in the data digging system of Web;
Figure 14 is provided by the present invention based on data-mining module figure in the data digging system of Web;
Figure 15 is provided by the present invention based on data mining explanation module figure in the data digging system of Web.
Embodiment
The present invention is further illustrated below in conjunction with accompanying drawing.
Each functional module of data digging system strictness based on Web provided by the present invention according to normal data excacation flow scheme design system, and make its each module independent, interfaceization, the variation of each inside modules can not cause the change that other module is big like this, such module independent helps the fast updating and the upgrading of system, to adapt to the develop rapidly of data mining technology.Concrete functional module as shown in Figure 3, wherein:
Data source connects the unified interface module: the connection of local data source, network data source and other data file is provided, generate the data object that mining algorithm can directly use, the interface of data object visit is provided, this functional module need realize the transparency of various data, it is converted into the defined data object of native system, thereby provides a kind of unified data object for processing for data mining algorithm.
Raw data visual presentation module: use performance complex data and mutual relationships thereof such as understandable point and line chart, histogram, pie chart, network diagramming, the user can be had comparatively intuitively raw data understand.
Data preprocessing module: provide raw data is carried out pre-service, comprise the various processing modes of data scrubbing, integrated, conversion and reduction; And each Preprocessing Algorithm is all carried out modular design, makes that adding new Preprocessing Algorithm is very easy to.
Interactively excavation module: this module provides the data mining algorithm of multiple employing different technologies, and provide the good man-machine interaction interface to various algorithms, the user can carry out the parameter setting and the modification of algorithm by it, utilize training data that algorithm model is trained, utilize test data that the mining model that obtains is assessed again, repeatedly repeatedly after, select a more excellent mining model of test result to carry out the data mining task that the user submits to.This module provides certain interactive capability, makes whole excavation, analytic process is controlled and easy to understand.Each algorithm of this module all adopts modular design, increases the independence of algoritic module and whole platform, improves the dirigibility and the extendability of total system.
Excavate the graphical display module of result: the function that this module is finished is to adopt various data visualizations and graphics technology, and the result of mining analysis is showed the user in the mode of figure.Native system provides intuitively, patterned mining analysis result, and the user can be understood easily.
The final goal of native system is the service that some data minings are provided on Internet, and the user carries out data mining analysis by the various service modules (as association analysis, cluster analysis, classification analysis etc.) that can use us to provide after registering to the data of oneself.This system other data digging system of comparing has its comparatively unique characteristics:
1) carries out data mining based on Web
The software and hardware investment that the client will carry out data mining, the business of analyzing is required is reduced to the IE browser, and the user need not to buy, any data mining software is installed, and just can obtain data mining service easily.When the user need carry out data mining analysis, only need data are submitted to data digging system provided by the present invention, according to the actual needs, select corresponding algorithm just can finish corresponding data mining task, the data mining that obtains wanting, analysis result.
2) unified data management
Native system can be realized the transparency of data object.Various data of different types all are converted to the uniform data object that the data mining algorithm of native system can use uniformly.No matter be the mining algorithm that adopts what digging technology like this, can both handle this data object, this has just realized the independence of algorithm and data.Just because of such function is arranged, could realize modularization, the independent of mining algorithm, can better adapt to the develop rapidly of data mining technology.It has also realized corresponding data preprocessing function, after having noise or other incomplete data through the data pre-service, just can become the data object that algorithm can be handled, and the various relatively mining algorithms of this preprocessing process, be fully transparent, independently.
3) the integrated storehouse of algorithm
Native system is realized modularization, the independent of mining algorithm by the algorithm general-purpose interface.Each algorithm is fully independently in the system, and we can add up-to-date algorithm at any time according to the actual needs, can not be some effects that good algorithm unloads easily also, these operations to the operate as normal of total system without any influence.Just because of it such function is arranged, we can provide the data mining service by the more integrated very easily algorithm than ripe, common data mining, also can allow theoretical research person easily in the test of the validity and the feasibility of the enterprising line algorithm of platform, the realization theory achievement is to the conversion at full speed of practical application rapidly.By the mode of algorithm tree, the general-purpose interface of algorithm is provided, follow these interfaces and all can seamlessly be integrated in the system according to the new algorithm that certain standard is write.
4) mining process of intelligent interaction
In native system, follow the process of data mining fully and come the definition of data mining task.In mining process, the user can instruct excavation by the man-machine interaction of height.The user can pass through the good man-machine interaction interface, carry out the setting of algorithm parameter easily, can realize algorithms of different model, mining model are assessed, utilize assessment result preferably mining model excavate service, thereby obtain mining analysis result preferably.
5) raw data is visual
Native system utilizes advanced data visualization technology, makes the user to have one to understand comparatively intuitively to data before mining analysis, helps further carrying out the data pre-service, also makes things convenient for the user better to carry out man-machine interaction when data mining.
6) the mining analysis result is graphical
Native system utilizes advanced visualization technique and graphics technology, shows the mining analysis result with graphing capability comparatively intuitively, helps the user like this and better understands the excavation result.
7) unified data-interface
Native system has been realized the transparency of various data sources, no matter be the relational database of what type, still general text or MS relevant documentation can seamlessly be connected with system.So just can expand the practicality of system greatly.
Native system is followed open industry standard, adopts the J2EE three-tier architecture to realize that this system is distributed, open, modular, and be general.The data of excavating may can be database or data warehouse or local data file from various data sources.So just need in the middle of various data sources and digging system, provide a translation interface, make digging system have unified Data View various data sources.Native system serves as main the design with the J2EE framework, wherein mainly comprises EJB server, 2 server components of Web server, and the Web layer adopts the Struts framework, and system architecture as shown in Figure 1.The EJB layer mainly provides the interface between Web service end and the EJB layer, and has realized the processing of data mining algorithm, and the main computing step of data mining analysis etc. all realizes in the EJB layer.The Web layer mainly provides the interface of and user interactions, accept the user input, provide mining process mutual and show the mining analysis result.The user realizes the mutual of user and mining process by browser access Web service end.
Because native system is based on Web the data mining service is provided, thereby the user utilizes browser and Web server to carry out alternately, so just can provide the data mining service for the user easily.All of native system and user mutual and represent content all undertaken by browser.
The Web service end is the request of being responsible for the response browser in the system architecture of native system, provides service for it, and provide interface with Enterprise Java Bean container, be used to represent html page and the part of accepting user's input, corresponding JSP wherein arranged, Servlet and JavaBean assembly to client.The pattern of MVC is followed in the Web layer strictness of MinerOnWeb system, and the process of handling with regard to system responses user's page request describes below:
1) controller Servlet receives the Http Request request from client, and is converted to the Event incident;
2) JavaBean that calls in the Model layer according to corresponding Event incident of controller Servlet begins to carry out business logic processing;
3) scheduling in the middle of the JavaBean in the Model layer realizes, the EJB assembly that calls in the Business Logic is realized service logic, the EJB assembly can be realized access and computing to business datum by database and algorithm controls engine;
4) controller Servlet is according to service processing result, and resolution path URL calls the corresponding JSP page;
5) correlation method in the JSP page invocation Business Logic is obtained data;
6) JSP generates html page according to data, returns browser, carries out page representation.
So just service logic, data exhibiting and the steering logic of Web layer are separated fully, such benefit is the extendability that has improved total system, is convenient to realize its modular function.In MinerOnWeb, the core missions of Web end are:
1) flow process of the whole data mining process of control.
2) provide user interface, accept user request, the input of interpreting user also is mapped as executable operation with them, and operation requests is passed to the EJB service end.
3) obtain data and the parsing that the EJB server end returns, data are shown to the user by the JSP page mode.
And the EJB service end mainly is made up of the EJB assembly, is used for the processing of service logic, is mainly used in the computing part of whole data mining processing and the interface and the processing capacity of input, exposition.The task of its core mainly contains:
1) data object and the algorithm object of management data excavation.
2) finish data mining capability interface and specific implementation, interface is offered the Web service end.
3) the data volume size in control and the Web service end reciprocal process.
For system function moduleization, the consideration of structure independent, adopt between module and the module as far as possible a unified interface carry out between the system data transfer and call, so just can realize the transparence between the module, the extensibility of enhanced system.Just consider that based on this some native system is introduced
Pattern realizes the communication of EJB service end and Web service end.The Facade pattern mainly contains following advantage as shown in Figure 5:
1) it shields subsystem components to the client, the function of using the method mode to provide subsystem to satisfy, thereby reduced the number of the object that the client handles and made subsystem more convenient to use;
2) realized that the loose coupling between subsystem and the client concerns, and the functional module of subsystem inside is tightly coupled often, the loose coupling relation makes the component variations of subsystem can not have influence on its client.The Facade pattern helps resume hierarchical structure system, also helps the dependence layering between the object;
3) if application need, it does not limit their and uses subsystem class, thereby can be selected between system's ease for use and versatility.
Only there is an EJB object in the Enterprise Java Bean container of native system EJB service end
, mainly utilize this EJB to finish interface function and control the data flows, and play managerial role.Native system is a multi-user's a system, need to keep the state of each user in mining process, so this EJB object is set to one state session Bean (Stateful Session Beans) is arranged, its effect is the control service logic, rather than finish concrete realization, there is state session Bean to keep the state that certain calls the client, and in diverse ways calls, keeps this state in man-to-man mode.In order to improve the efficient of system, Stateful Session Beans can be written into secondary storage devices (as hard disk) after free time certain client, after the client sends new call request, return to the internal memory from secondary storage devices again.Like this to improve response speed, save memory.
As shown in Figure 4, in the J2EE application server EJB object dispose with Web server in the mutual process of the Action of Struts framework and JSP be:
1) in the deployment phase, the EJB native object will be bundled on the JNDI tree of J2EE application server name service, and be assigned with a JNDI title;
2) Web server contacts JNDI name service for the first time to obtain the EJB native object;
3) Web server sends request by the Reference that uses the EJB native object to the EJB object;
4) the EJB native object is created (or searching) Facade EJB object;
5) local EJB object returns to client with the Reference of EJB object;
6) client obtains the Reference of EJB object, and calls correlation method in remote interface;
7) container is tackled calling of following method and it is appointed the example to Bean, and in fact the Bean example is forwarded to request and handles in the corresponding object, obtains rreturn value;
8) the EJB object by remote interface to client return results value.
Native system uses configuration file to realize by the management that EJB is configured file:
1) configuration information with system saves as file separately, needn't be hard-coded in the software, can reduce the complexity of system, increases the flexibility ratio of system configuration.When we increase algorithm, only need the change configuration file, the flow process of system and the judgement statement in the class not recompility program are changed, and just can finish, and seem simple and quick like this.
2) description of configuration file and form are very simple, and the user can just understand the whole process that increases algorithm by note, and it is clear understandable that total also seems.
Mainly contain two configuration files in the system now, be respectively DatabaseUtils.props data base configuration file and GenericObjectEditor.props algorithm configuration file, the transparent connection by managing these two configuration file fulfillment databases and the dynamic load of algorithm.
The entire process of native system is carried out in strict accordance with the data mining standard procedure, its concrete treatment scheme as shown in Figure 2, specific implementation process is as follows:
1) by after the upward long-range or local data source of data-interface connection, obtains corresponding data object;
2) by data visualization modules exhibit raw data, make the user impression intuitively be arranged to data;
3) utilize data preprocessing module that raw data is handled;
4) select suitable mining algorithm according to corresponding mining task;
5) respective algorithms is carried out the adjustment setting of some parameters;
6) utilize training data that corresponding algorithm model is trained, obtain the mining analysis model;
7) utilize test data that the mining analysis model is assessed, whether select the new algorithm model of needs, perhaps the parameter of original algorithm model is done corresponding adjustment according to assessment result;
8) utilizing the best mining model of assessment result in the above-mentioned steps that corresponding data object is carried out mining analysis handles;
9) the mining analysis result is graphically showed, make the user understand easily.
Below be specific embodiments more of the present invention:
The definition of buzzwords more of the present invention and foreign language initial group speech former as shown in the table on the same group:
| Sequence number | Term name | Explanation |
| 1. | Data mining | Data mining (Data Mining) is exactly identification or extract implicit, novel, the information of potentially useful and the process of knowledge from data a large amount of, incomplete, noisy, fuzzy, at random. |
| 2. | The data pre-service | According to the requirement of data mining algorithm, the data of selecting are carried out again operations such as projection, selection normalization conversion, so that the processing of mining algorithm. |
| 3. | Correlation rule | Association rule mining is exactly to excavate the relevant knowledge that connects each other between the valuable data of description item from lot of data. |
| 4. | Classification | Sorting technique is used for the discrete classification of predicted data object. |
| 5. | Cluster | Cluster is one data set is divided into the process of some groups or class, and makes the data object in same group have higher similarity; And the data object on the same group is not dissimilar. |
| 6. | J2EE | The architecture of the challenge that exploitation, deployment and the management that a kind of Java of utilization 2 platforms are simplified enterprise solution is relevant. |
| 7. | Struts | Struts is an Open Source project of Apache foundation Jakarta project team, and it adopts the MVC pattern, can help java developer to utilize J2EE exploitation Web to use well, and Struts also is an Object-Oriented Design simultaneously. |
| 8. | Weka | The Weka system mainly solves an open source software of data mining task in the real world by the exploitation of New Zealand waikato university with the algorithm of machine learning.Weka is a unit software, and the data mining research tool bag that function is more intense uses based on the demonstration as graphical interfaces of the GUI client of Swing. |
| 9. | Data object | Instantiation weka in the Instances class that defines. |
Some functional requirements of data digging system based on Web provided by the present invention are as follows;
The input of system: the source data that waits excavation
The output of system: the displaying (literal report, graphic result etc.) of excavating the result
Functional requirement (specifically as shown in Figure 6):
(1) authentication function;
(2) unified data management;
(3) data visualization function;
(4) data preprocessing function;
(5) data mining capability;
(6) explanation function as a result;
(7) intelligent interaction function.
Performance requirement: handle mass data, the computation complexity height.
Operation demand: mainly form by the data display program on foreground and the data computation program on backstage.System can timing automatic opens and finishes, and in the function that occurs can realizing under the unusual situation of some running environment opening and recovering automatically again automatically, promptly in the control of its operation artificial interference is reduced to than low degree.
The running environment of native system requires as follows:
Network environment: trunk is a Fast Ethernet, is adapted to Internet/Intranet.
Hardware environment: adopt high-performance, multi-processor computer system as the EJB server, handle the interrelated logic business, database server is positioned at any position of Internet.High-performance computer system provides page access as Web Application Server, and this distributed design can provide the computing power of magnanimity, guarantees data mining computation's reliability energy simultaneously.
Software environment: high-performance computer operating system, J2EE application server, Web Application Server.
Client: PC+web browser.
The principle of design of native system is as follows:
Data volume is big: data mining need be found out interested knowledge from mass data, so will take into full account data volume transmission, data computation problem in design process;
Expandability is strong: because mining algorithm is constantly expansion, so system must have good expandability.System design structuring as far as possible, modularization, and other subsystems are reserved corresponding interface.
Maintainable good: as to require system to have suitable dirigibility, so that safeguard.
Advanced: system adopts the open frame of international popular, and application software adopts the B/S structure, and Network Transmission adopts ICP/IP protocol.
The webpage flow process of native system designs the Web layer according to the Struts framed structure, the concrete operations process flow diagram as shown in Figure 7, the module logic diagram as shown in Figure 8, each functions of modules is described below table:
| The module title | The module numbering | Functional description | Control relation |
| Authentication module | G1 | Identifying user identity | |
| Initialization module | G2 | To system initialization, the different stage user is directed to the different pages | Controlled by G1 |
| The data source link block | G3 | The connection of various different data format data sources is provided, generates the data set that mining algorithm can directly use, the interface of data object visit is provided. | Controlled by G2 |
| Data visualization represents module | G4 | Use understandable point and line chart, histogram, pie chart, network diagramming, Interactive Visualization, dynamic similation, Computer Animated Graph performance complex data and mutual relationship thereof, make the user produce impression intuitively data. | |
| Data preprocessing module | G5 | Provide source data is carried out pre-service, comprise the various processing modes of data scrubbing, integrated, conversion and reduction; And be easy to increase new Preprocessing Algorithm. | Controlled by G2 |
| Excavate processing module | G6 | Polytype data mining algorithm processing mode is provided, carries out the data mining task that the user submits to.Visualized data interface and perfect interactive capability are provided, make whole excavation, analytic process is controlled and easy to understand. | Controlled by G2 |
| Excavate display module as a result | G7 | The function that this module is finished is that the result with data mining shows the user, provides intuitively, patterned excavation and analysis result, and the user can be understood easily. |
Provided by the present invention based on authentication module in the data digging system of Web as shown in Figure 9, initialization module as shown in figure 10, the data source link block as shown in figure 11, data visualization represents module as shown in figure 12, data preprocessing module as shown in figure 13; Data-mining module as shown in figure 14; The data mining explanation module as shown in figure 15.The realization of the every functional requirement of native system is as shown in the table with the relations of distribution of each piece program:
| G1 | G2 | G3 | G4 | G5 | G6 | G7 | |
| The authentication function | √ | ||||||
| Unified data management | √ | √ | √ | √ | |||
| The data visualization function | √ | √ | |||||
| The data preprocessing function | √ | ||||||
| Data mining capability | √ | ||||||
| Explanation function as a result | √ | ||||||
| The intelligent interaction function | √ | √ | √ | √ | √ |
The Interface design of the data digging system based on Web provided by the present invention is as follows:
External interface:
1, EJB and web alternately as shown in Figure 4.In the J2EE application server EJB object generate with Web server in the mutual process of the Action of Struts framework and JSP be:
1) in the deployment phase, the EJB native object will be bundled on the JNDI tree of J2EE application server name service, and be assigned with a JNDI title.
2) Web server contacts JNDI name service for the first time to obtain the EJB native object.
3) Web server sends request by the Reference that uses the EJB native object to the EJB object.
4) the EJB native object is created (or searching) Facade EJB object.
5) local EJB object returns to client with the Reference of EJB object.
6) client obtains the Reference of EJB object, and calls correlation method in remote interface.
7) container is tackled calling of following method and it is appointed the example to Bean.In fact the Bean example is forwarded to request in the corresponding object and handles, and obtains rreturn value.
8-9) the EJB object by remote interface to client return results value.
Facade EJB is the key that realizes systemic-function.When we the EJB service end jar file be published on the application server, Web server just can far call Facade, realizes function corresponding.
2, the connection of database
Method by JDBC connects database, the type of database difference, and the driver that needs will be different.Use the mode stored data base configuration of configuration file.
Internal interface:
EJB internal interface (class empty expression by name is arranged in same class with a last interface, and following each epiphase together)
| Sequence number | Interface name | Definition | |
| 1. | ?connectDatabase | ?public?void ?connectDatabase(String?URL， ?String?login，String?password， ?String?query) | Connect database |
| 2. | ?getArithmeticCatego ?ry | ?public?String ?getArithmeticCategory() | Obtain algorithm classification |
| 3. | ?getArithmeticInfo | public?Properties getArithmeticInfo() | Obtaining parameter information and value is by resolving the character string of concrete characterising parameter information |
| 4. | ?getAttributesName | public?String getAttributesName(int?index) | Obtain the data recording Property Name |
| 5. | ?getAttributesType | public?String getAttributesType(int?index) | Obtain the data recording attribute type |
| 6. | ?getAttributesValues | public?Object[][] getAttributesValues(int?index) | Obtain the occurrence of data recording attribute |
| 7. | ?getClassIndex | public?int?getClassIndex() | |
| 8. | ?getDistinctCount | public?int?getDistinctCount(int index) | Obtain the number of different attribute |
| 9. | ?getFilelnstances | public?void getFileInstances(String filesource，long?flength) | From data file, obtain object data set |
| 10. | ?getInAttribute | public?Attribute getInAttribute(int?index) | Obtain the attribute of data set |
| 11. | ?getInAttributeStats | public?AttributeStats getInAttributeStats(int?index) | Obtain the statistical information of data set attribute |
| 12. | ?getInAValue | public?double?getInAValue(int m，intn) | Obtain the value that the m bar writes down n attribute |
| 13. | ?getInstanceValue | public?double getInstanceValue(int instanceIndex，int?attributeIndex) | Obtain value (the another kind of method of certain attribute of certain record?) |
| 14. | ?getIsMissing | public?boolean?getIsMissing(int instanceIndex，int?attributeIndex) | The value of judging certain attribute of certain record is to lose |
| 15. | ?getM_arithmetic | public?Object getM_arithmetic() | Obtain the current algorithm object |
| 16. | ?getMax | public?double?getMax(int?index) | Obtain the numerical value the maximum in the statistics of attributes |
| 17. | ?getMin | public?double?getMin(int?index) | Obtain the numerical value reckling in the statistics of attributes |
| 18. | ?getMissingCount | public?int?getMissingCount(int index) | Obtain the attribute number that property value is lost |
| 19. | ?getNominalCounts | public?int[] getNominalCounts(int?index) | Obtaining attribute type is the number of Nominal |
| 20. | ?getNominalLength | public?int?getNominalLength(int index) | |
| 21. | ?getNumAttributes | public?int?getNumAttributes() | Return all properties number |
| 22. | ?getNumericCount | public?double getNumericCount(int?index) | Obtaining attribute type is the number of Numeric |
| 23. | ?getNumInstances | public?int?getNumInstances() | The bar number of return data record |
| 24. | ?getNumValues | public?int?getNumValues(int index) |
| 25. | getRelationName | public?String?getRelationName() | Obtain dataset name |
| 26. | getStaticInfoOflnsta nces | public?String getStaticInfoOflnstances() | Obtain the statistical information of data recording |
| 27. | getStdDev | public?double?getStdDev(int index) | |
| 28. | getTotalCount | public?int?getTotalCount(int index) | |
| 29. | getUniqueCount | public?int?getUniqueCount(int index) | |
| 30. | setArithmeticParame ters | public?Boolean setArithmeticParameters(int paraID，String?proValue， Int[]tagID，String[]descriptions) | SelectedTag type in the processing parameter |
| 31. | setArithmeticParame ters | Public?Boolean setArithmeticParameters(int paraID，String?proValue ) | Simple types in the processing parameter and some other type |
| 32. | setM_arithmetic | public?void setM_arithmetic(String arithmeticName) | Select the selection of face already, a newly-generated algorithm object in facade according to the user at algorithm |
| 33. | setParaSetter | public?Void?setParaSetter() | The m_Setters and the m_Types of parameter are set |
| 34. | startArithmetic | public?String?startArithmetic() | Carry out the algorithm computing |
| 35. | startFilterAlgr | public?Void?startFilterAlgr(int classIndex) | Carry out the Preprocessing Algorithm computing |
| 36. |
The mode of the operation control of native system: the Web service end belongs to key-course and view layer, and its workflow is: display page, accept and resolve user's input, and the process user request passes to the EJB service end with associative operation, and return results is shown to the user.The EJB service end belongs to model layer, is responsible for the realization of concrete business function.Realize data storage, data mining computing, mining algorithm parameter concrete business is set etc., and provide the correlation function interface.
Operation steps:
1, the address of input MinerOnWeb will enter the Login.jsp page, and the prompting user imports username and password.
2, press and determine that will move LoginAction behind the key verifies username and password, if be proved to be successful, then change choice.jsp over to, authentication failed then changes the error.jsp page over to.
3, can select the Data Source that is used to excavate in choice.jsp, can be local data file or the tables of data in the database at present.If the selection local data file has just entered fileLoad.jsp, press and browse, fileLoadingAction can select the data file that will upload with operation, then by " upload " button, upload file.Enter doupload.jsp, the relevant information (path of filename, file size, file extension, upload file) of showing file, point submit submits to, operation getFileAction, by the remote interface of JNDI acquisition Facade EJB, the data in EJB remote method invocation (RMI) acquisition resolution file.Obtain the summary information of data set by the EJB remote interface.
If select to connect database, then change the connectDB.jsp page over to, fill in database address, user name, password etc., select present attachable type of database, determine that the back system will move connectDBAction, parameter according to Query Database connects database by JDBC, enter showTable.jsp, operation getTableAction shows the tables of data in the database, select single or a plurality of tables of data, after the submission, change showTablesDetail.jsp over to, the showInfo.jsp page, in the page, use Applet, with data presentation by parameter.The user can check the statistical information of data set, if any how many bar records, and how many attributes, the spans of each attribute etc. also can be by carrying out next step algorithm computing at this attribute of need selecting.
4, after the user puts next page, operation showInfoAction reads configuration file by the EJB remote interface, obtains the now available mining algorithm set of MinerOnWeb, changes AlgrmResult.jsp over to, and the user can select algorithm wherein in the page.
5, behind the selection algorithm, will move AlgrmResultAction.According to the algorithm title of selecting, same is the parameter value that obtains algorithm by remote interface, changes Paramenters.jsp and demonstration over to.The user can dispose each parameter value as required, calculates so that carry out data mining.
6, in Paramenters.jsp, when the user click determine after, operation ParametersAction, the purpose of this Action is by remote interface, and the parameter that user's modification is crossed is passed to Facade EJB, revises algorithm image parameter among the EJB by Facade.Then, call the data mining algorithm of EJB end training data is handled, obtain corresponding mining model.Utilize test data that mining model is assessed again, the different parameters of algorithm can also repeatedly be set or select no algorithm to obtain different mining models, utilize assessment result preferably mining model carry out corresponding data mining task, after obtaining the result of mining analysis, change the result.jsp page over to.The mining analysis result is presented in the result.jsp page, realizes mining analysis result's visual presentation.
Main JSP and Action that native system used are described below:
Login.jsp: the portal page of system, in this page, import the user name and password.
LoginAction.java: whether the username and password of judging input has enough authorities to enter next step operation, if be proved to be successful, initialization EJB object Facade obtains the availability database type and enters connectDB.jsp.
ConnectDB.jsp: the page that configuration database connects, input database address, user name, password and query sentence of database in this page.
ConnectDBAction.java: the connection database information of input passes to EJB object remote interface, connects database by the EJB end, generates data object.
ShowInfo.jsp: the concise and to the point statistical information of video data object.
ShowInfoAction.java: the remote interface that provides by the EJB object obtains algorithm list.
AlgrmResult.jsp: the page of display algorithm inventory.The algorithm that the user need can select on this page comes the data object is handled.
GetBeanInfoAction.java: according to selected algorithm, instantiation algorithm object in EJB, and the parameter list of acquisition algorithm object.
Parameters.jsp: show the modifiable parameter value of this algorithm, the user can revise parameter in the page.
SetPrametersAction.java: the parameter after the modification value is passed to EJB, call the parameter value of revising algorithm object among the EJB by Method.In EJB, carry out the computing of data mining.
Result.jsp: display algorithm result's the page.
ResultAction.java: turn back to the Login.jsp page again.
Crror.jsp: the page that expression makes mistakes.When mistake appears in system, jump to this page.
Claims (10)
1, a kind of data digging system based on Web, it mainly comprises EJB server, Web server and database, wherein the EJB server provides the interface between Web service end and the EJB layer, and realized the processing of data mining algorithm, the Web layer provides the interface of and user interactions, accept the user input, the mining process interactive interface is provided and shows the mining analysis result, it is characterized in that comprising following module:
Authentication module: is that the user carries out authentication by Internet to browser;
Initialization module:, the user of different stage is directed to the different pages to system initialization;
Data link block: the connection of various different data format data sources is provided, generates the database that mining algorithm can directly use, the interface of data object visit is provided;
Data visualization represents module: use the understandable media performance complex data and the correlationship thereof that can produce visual impression;
Data preprocessing module: provide source data is carried out pre-service, comprise the various processing modes of data scrubbing, integrated, conversion and reduction;
Mining model evaluation module: utilize different mining models that test data is carried out mining analysis, the Different Results that obtains is assessed, select corresponding mining model to carry out data mining according to assessment result;
Excavate processing module: the data mining algorithm processing mode is provided, carries out the data mining task that the user submits to;
Excavate display module as a result: the result of data mining is showed the user by media, provide intuitively, patterned excavation and analysis result.
2, the data digging system based on Web according to claim 1 is characterized in that, EJB service end and Web service end are linked up by the Facade pattern.
3, the data digging system based on Web according to claim 1 and 2, it is characterized in that, the Web service end be responsible for responding browser request, provide service for it, interface with the EJB server is provided, be used to represent html page and the importation of accepting the user, comprise JSP, Servlet and JavaBean assembly to client.
4, the data digging system based on Web according to claim 1, it is characterized in that described data visualization represents module and excavates media in the display module as a result and can be in point and line chart, histogram, pie chart, network diagramming, Interactive Visualization, dynamic similation, the computer animation one or several.
5, the data digging system based on Web according to claim 1 is characterized in that, described data preprocessing module and excavation processing module are provided with the interface that can increase new data algorithm at any time.
6, the data digging system based on Web according to claim 1, it is characterized in that, database can be arranged on the server of Internet optional position, drives being connected of WEB end assembly and database by JDBC, and supports the data file of client to upload to WEB and hold assembly; WEB end assembly obtains Data Post and becomes corresponding data object, and this data object is passed to the processing of applied business logic module; The WEB assembly can be placed on respectively on the different servers with the applied business logic module, also can be on same server, and both realize remote object invocation by RMI-IIOP mechanism; The data link block adopts multi-thread mechanism, and keeps synchronously.
7, the data digging system based on Web according to claim 1 is characterized in that, system follows Struts1.1 framework bag and EJB2.0 standard based at the J2EE of Java 2 platform, adopts MVC three-layer architecture model; Business Logic mainly comprises EJB object, JavaBean; Key-course mainly comprises the Action action class based on Servlet; The view layer mainly is made of the JSP page; Application server and WEB server can adopt various standard J2EE application servers; Wherein WEB end assembly can run on various computer systems; The application end assembly can run on various computer systems.
8, a kind of control method of the data digging system based on Web is characterized in that, comprises following steps:
(1), connect long-range or local data source, obtain corresponding data object by data-interface;
(2), show raw data by the data visualization module to remote user end;
(3), utilize data preprocessing module that raw data is handled;
(4), select suitable mining algorithm according to mutually deserved mining task;
(5), respective algorithms is carried out the parameter adjustment setting and reach suitable algorithm adjustment;
(6), utilize training data that corresponding algorithm model is trained, obtain the mining analysis model;
(7), utilize test data that the mining analysis model is assessed, whether select the new algorithm model of needs according to assessment result, perhaps the parameter of original algorithm model is done corresponding adjustment;
(8), utilizing the best mining model of the middle assessment result of step (7) that corresponding data object is carried out mining analysis handles;
(9), will excavate the result utilizes the medium and holds displaying to long-range or local user.
9, the control method of the data digging system based on Web according to claim 8 is characterized in that, all of native system and user mutual and represent content all undertaken by browser.
10, the control method of the data digging system based on Web according to claim 8 is characterized in that, the step that system responses user's page request is handled is as follows:
(1), controller Servlet receives the Http Request request from client, and is converted to the Event incident;
(2), the JavaBean that calls in the Model layer according to corresponding Event incident of controller Servlet begins to carry out business logic processing;
(3), the JavaBean in the Model layer realize in the middle of scheduling, the EJB assembly that calls in the Business Logic is realized service logic, the EJB assembly can be realized access and computing to business datum by database and algorithm controls engine;
(4), controller Servlet is according to service processing result, resolution path URL calls the corresponding JSP page;
(5), the correlation method in the JSP page invocation Business Logic is obtained data;
(6), JSP is according to data, generates html page, returns browser, carries out page representation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100226681A CN100476819C (en) | 2006-12-27 | 2006-12-27 | Data mining system based on Web and control method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100226681A CN100476819C (en) | 2006-12-27 | 2006-12-27 | Data mining system based on Web and control method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1975720A CN1975720A (en) | 2007-06-06 |
| CN100476819C true CN100476819C (en) | 2009-04-08 |
Family
ID=38125790
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2006100226681A Expired - Fee Related CN100476819C (en) | 2006-12-27 | 2006-12-27 | Data mining system based on Web and control method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100476819C (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101908191A (en) * | 2010-08-03 | 2010-12-08 | 深圳市她秀时尚电子商务有限公司 | Data analysis method and system for e-commerce |
| CN105357027A (en) * | 2015-09-24 | 2016-02-24 | 四川长虹电器股份有限公司 | Lightweight data service bus system based on large data |
| CN103853821B (en) * | 2014-02-21 | 2017-02-22 | 河海大学 | Method for constructing multiuser collaboration oriented data mining platform |
Families Citing this family (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101231661B (en) * | 2008-02-19 | 2010-06-23 | 上海估家网络科技有限公司 | Method and system for digging object grade knowledge |
| CN101876896B (en) * | 2009-04-30 | 2013-04-03 | 深圳市永兴元科技有限公司 | E-government affair development system for promoting informationization technology |
| CN102054001B (en) * | 2009-10-28 | 2012-10-03 | 中国移动通信集团公司 | Data preprocessing method, system and device in data mining system |
| CN101968812A (en) * | 2010-10-25 | 2011-02-09 | 中国农业大学 | Method and device for automatically generating cartogram by calling database data |
| CN103020006B (en) * | 2011-09-24 | 2016-09-07 | 国家电网公司 | A kind of equipment state prediction method excavated based on mass data |
| CN103309867A (en) * | 2012-03-09 | 2013-09-18 | 句容智恒安全设备有限公司 | Web data mining system on basis of Hadoop platform |
| CN103425707A (en) * | 2012-05-25 | 2013-12-04 | 中兴通讯股份有限公司 | Data analyzing method and data analyzing device |
| CN103136337B (en) * | 2013-02-01 | 2016-05-04 | 北京邮电大学 | For distributed knowledge data mining device and the method for digging of complex network |
| CN104346376B (en) * | 2013-07-31 | 2017-11-03 | 红有软件股份有限公司 | Method and system of the data mining algorithm dynamic insertion to data mining platform |
| CN104166701B (en) * | 2014-08-04 | 2018-07-31 | 深圳先进技术研究院 | machine learning method and system |
| CN104408294B (en) * | 2014-10-31 | 2017-07-21 | 南京理工大学 | A kind of event summary method based on event relation network |
| CN104572074B (en) * | 2014-12-08 | 2019-04-05 | 北京辰闰丰青信息技术有限公司 | Based on big data graphical representation custom-built system |
| CN104537001A (en) * | 2014-12-15 | 2015-04-22 | 中国石油天然气股份有限公司 | Platform and method for mining oil gas information data |
| CN105045931A (en) * | 2015-09-02 | 2015-11-11 | 南京邮电大学 | Video recommendation method and system based on Web mining |
| CN106372240B (en) * | 2016-09-14 | 2019-12-10 | 北京搜狐新动力信息技术有限公司 | Data analysis method and device |
| CN106503039A (en) * | 2016-09-20 | 2017-03-15 | 南京邮电大学 | A kind of visualization real time data digging system and method |
| CN106528682A (en) * | 2016-10-25 | 2017-03-22 | 安徽讯呼信息科技有限公司 | Big-data text mining system of call center |
| CN106484914A (en) * | 2016-10-26 | 2017-03-08 | 国云科技股份有限公司 | A kind of modular assembly method for quickly realizing data mining analysis |
| CN106599325A (en) * | 2017-01-18 | 2017-04-26 | 河海大学 | Method for constructing data mining visualization platform based on R and HighCharts |
| CN107025288A (en) * | 2017-04-14 | 2017-08-08 | 四川九鼎瑞信软件开发有限公司 | Distributed data digging method and system |
| CN107145435A (en) * | 2017-05-27 | 2017-09-08 | 北京仿真中心 | A kind of assessment of performance system and method based on B/S frameworks |
| CN109388661B (en) | 2017-08-02 | 2020-04-21 | 创新先进技术有限公司 | Model training method and device based on shared data |
| CN108133734A (en) * | 2017-12-21 | 2018-06-08 | 广东工业大学 | A kind of analysis method, device and the equipment of medical expense big data |
| CN108829704A (en) * | 2018-04-28 | 2018-11-16 | 安徽瑞来宝信息科技有限公司 | A kind of big data distributed libray Analysis Service technology |
| CN108897587B (en) * | 2018-06-22 | 2021-11-12 | 北京优特捷信息技术有限公司 | Pluggable machine learning algorithm operation method and device and readable storage medium |
| CN109558395A (en) * | 2018-10-17 | 2019-04-02 | 中国光大银行股份有限公司 | Data processing system and data digging method |
| CN110633308A (en) * | 2019-08-28 | 2019-12-31 | 北京浪潮数据技术有限公司 | Data mining method, system and related device |
| CN111260969B (en) * | 2020-03-06 | 2021-12-14 | 华南农业大学 | Data mining course teaching practice system and teaching practice method based on system |
| CN113515506A (en) * | 2020-04-10 | 2021-10-19 | 中国石油化工股份有限公司 | LDAR (laser direct reflectance assessment) system and method based on big data mining analysis |
| CN112632146B (en) * | 2020-12-03 | 2023-04-07 | 成都大数据产业技术研究院有限公司 | Multi-person collaborative visual data mining system |
| CN112486475A (en) * | 2020-12-03 | 2021-03-12 | 成都大数据产业技术研究院有限公司 | Visual business modeling and model management system based on big data platform |
| CN112667702A (en) * | 2020-12-03 | 2021-04-16 | 成都大数据产业技术研究院有限公司 | Big data-based data mining system |
| CN112508667A (en) * | 2020-12-08 | 2021-03-16 | 安徽经邦软件技术有限公司 | Financial data analysis system based on cloud native micro-service architecture |
| CN112596853A (en) * | 2020-12-08 | 2021-04-02 | 青岛积成电子股份有限公司 | Method for realizing pluggable artificial intelligence visualization |
| CN114896477A (en) * | 2022-06-08 | 2022-08-12 | 徐州医科大学 | Data mining safety visualization system and method supporting multiple language algorithms |
-
2006
- 2006-12-27 CN CNB2006100226681A patent/CN100476819C/en not_active Expired - Fee Related
Non-Patent Citations (6)
| Title |
|---|
| 一种基于J2EE与XML的虚拟企业数据挖掘体系结构. 王名茗，王卫平，俞栋.计算机工程与应用，第2005年第4期. 2005 |
| 一种基于J2EE与XML的虚拟企业数据挖掘体系结构. 王名茗，王卫平，俞栋.计算机工程与应用，第2005年第4期. 2005 * |
| 基于J2EE的Web挖掘原型系统的研究与应用. 贾宇波，王永利，孙淑荣.计算机应用研究，第2003年第4期. 2003 |
| 基于J2EE的Web挖掘原型系统的研究与应用. 贾宇波，王永利，孙淑荣.计算机应用研究，第2003年第4期. 2003 * |
| 基于J2EE的空间数据挖掘系统设计与实现. 涂建东，陈崇成，黄洪宇，张群洪.计算机应用，第25卷第3期. 2005 |
| 基于J2EE的空间数据挖掘系统设计与实现. 涂建东，陈崇成，黄洪宇，张群洪.计算机应用，第25卷第3期. 2005 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101908191A (en) * | 2010-08-03 | 2010-12-08 | 深圳市她秀时尚电子商务有限公司 | Data analysis method and system for e-commerce |
| CN103853821B (en) * | 2014-02-21 | 2017-02-22 | 河海大学 | Method for constructing multiuser collaboration oriented data mining platform |
| CN105357027A (en) * | 2015-09-24 | 2016-02-24 | 四川长虹电器股份有限公司 | Lightweight data service bus system based on large data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1975720A (en) | 2007-06-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100476819C (en) | Data mining system based on Web and control method thereof | |
| CN110989983B (en) | Zero-coding application software rapid construction system | |
| CN1713196B (en) | Product ordering system based on automatic design grid | |
| CN101477572B (en) | Method and system of dynamic data base based on TDS transition data storage technology | |
| Bara et al. | A model for business intelligence systems' development | |
| CN103019673A (en) | Intelligent decision-making and entity recommending union system based on internet and work flow | |
| Boudriga et al. | Intelligent agents on the web: A review | |
| CN107103064A (en) | Data statistical approach and device | |
| CN111813958A (en) | Intelligent service method and system based on innovation and entrepreneurship platform | |
| Aghimien et al. | A review of the application of data mining for sustainable construction in Nigeria | |
| CN106940724B (en) | Multi-paradigm fusion analysis processing method for big data | |
| Huhns et al. | All agents are not created equal | |
| Burstein et al. | Model management and solvers for decision support | |
| Yang | JAVA Architecture of Chinese Online Guiding Systematic Framework based on Data Mining and Artificial Intelligence | |
| Wiederhold | Information systems that really support decision-making | |
| Zhang et al. | Intelligent business cloud service platform based on SpringBoot framework | |
| Qi et al. | Optimization design and implementation of shared information management system for industrial design network platform | |
| Smirnov et al. | Linked-data integration for workflow-based computational experiments | |
| Raza et al. | BIG DATA V’S MODELS, CHALLENGES, HADOOP ECOSYSTEM, ISSUES, USES, BENEFITS AND APPLICATIONS | |
| Jansen | Exploring interactive application landscape visualizations based on low-code automation | |
| Vargas-Acosta et al. | Towards SWIM Narratives for Sustainable Water Management. | |
| Li et al. | Developing an e-commerce application by using content component model | |
| Peristeras et al. | Reengineering Public Administration through Semantic Technologies and a Reference Domain Ontology. | |
| Natalija et al. | ENVIRONMENTAL RISK ANALYSIS WITH APPLICATION OF INTELLIGENT GEOGRAPHIC INFORMATION SYSTEM | |
| Xue | Design and Implementation of Enterprise Management Intelligent Decision System Based on Data Analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090408 Termination date: 20111227 |