CN101620597A - Method for analyzing product association of data service in mobile communication industry - Google Patents
Method for analyzing product association of data service in mobile communication industry Download PDFInfo
- Publication number
- CN101620597A CN101620597A CN200810039885A CN200810039885A CN101620597A CN 101620597 A CN101620597 A CN 101620597A CN 200810039885 A CN200810039885 A CN 200810039885A CN 200810039885 A CN200810039885 A CN 200810039885A CN 101620597 A CN101620597 A CN 101620597A
- Authority
- CN
- China
- Prior art keywords
- data
- model
- mobile communication
- service
- correlation rule
- 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.)
- Pending
Links
Images
Landscapes
- Telephonic Communication Services (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a method for analyzing the product association of a data service in the mobile communication industry, which comprises the following steps of: building a data set; building a plurality of association rule models; evaluating and testing the plurality of association rule models; selecting the optimal association rule model; and releasing and applying the model. Compared with the prior art, the method can well mine the association rule models aiming at a mobile communication service.
Description
Technical field
The present invention relates to data mining technology, particularly relate to the method for analyzing product association of data service in mobile communication industry.
Background technology
Segment market, make every effort to provide the epoch of differentiated services at operation enterprise, what domestic telecommunication operator faced is the market competitive pressure that increases day by day, can have sensitive, the powerful rapidly analytic system of a cover, formulate policy targetedly, take out various business strategys, it is more and more important to seem in good time.In recent years, domestic telecommunication operator had introduced data warehouse and data digging system at CRM and charge system in succession, and had obtained success.
Data mining (Data Mining) is exactly the interested knowledge of extracting data people from large database.These knowledge are information unknown, potentially useful that imply, prior, and the representation of knowledge of extraction is notion (Concepts), rule (Rules), rule (Regularities), pattern forms such as (Patterns).Data mining utilizes various analysis tools to find relation between model and data in mass data, and these models and relation can be used for giving a forecast.Data mining is by prediction future trend and behavior, makes prediction, based on the decision-making of knowledge.
The discovery of correlation rule is the most successful and most important task in the data mining, and its target is to find all frequent modes of data centralization.Mining Association Rules is the process in one two step:
(1) finds out Frequent Item Sets all among the transaction database D according to minimum support.
(2) produce strong correlation rule by Frequent Item Sets and min confidence, also can use additional interest-degree to come rule is measured.The basic model of association rule mining as shown in Figure 1.
Among Fig. 1, D is the transaction database of input, and the first step is found out Frequent Item Sets according to searching algorithm, and second step concentrated from frequent item and produces useful correlation rule, the correlation rule set that output is at last excavated.The user can carry out alternately with two sub-steps respectively by specifying minimum support (minsup) and lowest confidence (minconf), and by with alternately the result being made an explanation and assessing of output set.
Abroad, association rule mining has dropped into application, and domestic research and the regular optimization of tolerance that mainly concentrates on algorithm itself up to now, does not propose a kind of at systematized flow process mobile communication business, association rule mining, real-time automated method as yet.
Summary of the invention
Technical matters to be solved by this invention is exactly the method for analyzing product association that data service in mobile communication industry is provided for the defective that overcomes above-mentioned prior art existence.
Purpose of the present invention can be achieved through the following technical solutions: the method for analyzing product association of data service in mobile communication industry, it is characterized in that, and may further comprise the steps:
(1). according to the mobile communication business demand, make up the business datum fairground of service-oriented theme from the mobile communication business data warehouse;
(2). based on the business datum fairground, call Data Mining Tools, selected correlation rule training dataset makes up a plurality of correlation rule models;
(3). above-mentioned a plurality of correlation rule models are assessed and checked, select optimum correlation rule model, and this model is write model bank;
(4). the model in issue, the application model storehouse.
Described step (1) further comprises:
(11). according to the mobile communication business demand, the professional theme that specified data is excavated;
(12). select professional theme corresponding service detail list and related data dimension table;
(13). associated services detail list and related data dimension table;
(14). the data after the association are carried out pre-service, obtain the business datum fairground.
Described step (3) further comprises:
(31). according to the assessment inspection parameter of each correlation rule model, select optimum correlation rule model;
(32). invoke script writes model bank in real time with optimum correlation rule model.
Described mobile communication business comprises cell phone television services, mobile Ring Back Tone service, mobile ring service, moves complete bent business.
Described pre-service comprises:
Business datum in the related data dimension table is cleaned, filtered heavily;
With the format conversion of cleaning, filtering the business datum after weighing is the data layout that the correlation rule model is suitable for.
Described Data Mart comprises pretreated day business datum, month business datum and all business datums.
Compared with prior art, the present invention can carry out the excavation of correlation rule model well at mobile communication business.
Description of drawings
Fig. 1 is the ultimate principle of association rule mining of the present invention;
Fig. 2 is a process flow diagram of the present invention.
Embodiment
The invention will be further described below in conjunction with accompanying drawing.
As shown in Figure 2, the method for analyzing product association of data service in mobile communication industry may further comprise the steps:
(1). according to the mobile communication business demand, make up the business datum fairground of service-oriented theme from the mobile communication business data warehouse;
(2). based on the business datum fairground, call Data Mining Tools, selected correlation rule training dataset makes up a plurality of correlation rule models;
(3). above-mentioned a plurality of correlation rule models are assessed and checked, select optimum correlation rule model, and this model is write model bank;
(4). the model in issue, the application model storehouse.
Described step (1) further comprises: according to the mobile communication business demand, and the professional theme that specified data is excavated; Select professional theme corresponding service detail list and related data dimension table; Associated services detail list and related data dimension table; Data after the association are carried out pre-service, obtain the business datum fairground.
Described step (3) further comprises: according to the assessment inspection parameter of each correlation rule model, select optimum correlation rule model; Invoke script writes model bank in real time with optimum correlation rule model;
Described mobile communication business comprises cell phone television services, mobile Ring Back Tone service, mobile ring service, moves complete bent business;
Described pre-service comprises: the business datum in the related data dimension table is cleaned, filtered heavily; With the format conversion of cleaning, filtering the business datum after weighing is the data layout that the correlation rule model is suitable for;
Described Data Mart comprises pretreated day business datum, month business datum and all business datums.
The present invention selects in the world method in common opinion CRISP-DM (inter-trade data mining normal stream) as instructing, the method for analyzing product association of data service in mobile communication industry is proposed, by this flow and method, connect data warehouse, dispatch seamless integrated Data Mining Tools, the mobile communication business data are analyzed, find out the significant correlation rule between the data set discipline, carry out assessment, issue and the application of model, thereby provide decision support for next step marketing activity.But whole process timer-triggered scheduler is carried out, and full automation is finished the renewal and the issue of model.
Mobile communication business specific to association rule mining, specifically comprises cell phone television services, mobile Ring Back Tone service, mobile ring service, moves complete bent business; Mobile phone TV services according to the user capture mode, are divided into live, non-live, program request, download, and wherein non-live data is the union of program request and download.
Making up the Data Mart of service-oriented theme, is the demand according to mobile communication business, connects data warehouse, obtains professional detail list and relevant dimension table, carries out the pre-service of data, obtains correlation rule desired data collection; The data pre-service is meant related, cleaning, heavy, the format conversion of filter to data traffic table and dimension table; The data that the correlation rule model uses can be transaction form or table format,, select the transaction form for use herein; Model writes model bank, is meant to call write-in program, and the correlation rule model is write in the model bank in real time, and the composition of each rule of model comprises: the degree of confidence of the consequence of the cause of rule, rule, the matching degree of rule, rule, the most relevance number of rule; Application of model is meant model is applied on the target data set that the result is applied.
The issue of described model and model application result is meant the issue of model itself, the issue of model application result on certain data set, and the user can download model and model application result.
The detailed process of method of the present invention:
(1) Data Mart of structure service-oriented theme
The professional theme that specified data is excavated: understand business demand, the theme that specified data is excavated;
Select data: professional detail list and the related data dimension table of selecting professional theme;
Related fact table and Wei Biao: the professional detail list (fact table) of professional theme is associated with relevant dimension table.
The data pre-service: data cleansing, filter weigh, eliminate noise, derive and calculate the missing value data, the related data data volume of mobile service is very big, the renewal frequency height, data on each time scale have reflected different reality, different information are provided, at this,, prepare the data of three time scales: the moon, week, day specific to each professional theme; When extracting the data of three time scales, need filter heavily, eliminate duplicate record data, for example, the same TV programme of certain user capture repeatedly, corresponding have many records at lane database, after filter is heavy, merges into a record; The translation data form, with the data layout that data-switching becomes the correlation rule model to use, the data that the correlation rule model uses can be transaction form or table format,, select the transaction form for use herein.
(2) make up the correlation rule model
The Data Mining Tools that application seamless is integrated, based on selected correlation rule training dataset, the parametric optimization of realization business and data-driven makes up the correlation rule model.
Find out Frequent Item Sets: find out Transaction Information according to minimum support and concentrate all Frequent Item Sets;
Generation rule: produce strong correlation rule by Frequent Item Sets and min confidence.
(3) model evaluation and preferred
Model evaluation: parameter such as, degree of confidence rate poor according to degree of confidence, select optimization model;
The model warehouse-in: invoke script (write-in program), the correlation rule model that generates is write in the model bank in real time, the particular content of model comprises: the degree of confidence of the consequence of the cause of rule, rule, the matching degree of rule, rule, the most relevance number of rule.
(4) issue of model and application
The model issue: at the foreground of system issue correlation rule model, the user can see the details of model strictly all rules, comprises cause, consequence, matching degree, degree of confidence, most relevance number;
Model is used: select a certain rule, the select target data set to the data set application rule, finds data centralization to have cause but does not have all records of consequence;
The issue of model application result: at the application result of system's foreground real-time release correlation rule, the user can check and the down load application result.
Embodiment
This sentences certain commmunication company's mobile TV program association analysis, specifically selects live broadcast service is example, finishes whole association rule mining process.
(1) professional understanding and data are selected
Certain commmunication company's mobile phone television living broadcast business is carried out the association analysis of program herein, select the professional detail list and the related data dimension table of professional theme.
(2) data processing
The detail list of professional theme is coupled together with relevant dimension table,, obtain the data of the transaction form of the moon, week, day three time scales by heavy, the format conversion of cleaning, filter of data.
Result is Time of Day yardstick tables of data (DM_AR_DAY_ZB), time-of-week yardstick tables of data DM_AR_WEEK_ZB, month yardstick tables of data DM_AR_MONTH_ZB.
The field of three tables is identical, list of fields such as table 1:
Table 1
(3) generate the correlation rule model
Obtain model result such as table 2 according to shaku degrees of data collection:
Table 2
Model summary information such as table 3:
Table 3
(4) model is used
The rule of option table 4:
Table 4
The preference pattern application data set, application rule, analysis result table 5:
Table 5
Claims (6)
1. the method for analyzing product association of data service in mobile communication industry is characterized in that, may further comprise the steps:
(1). according to the mobile communication business demand, make up the business datum fairground of service-oriented theme from the mobile communication business data warehouse;
(2). based on the business datum fairground, call Data Mining Tools, selected correlation rule training dataset makes up a plurality of correlation rule models;
(3). above-mentioned a plurality of correlation rule models are assessed and checked, select optimum correlation rule model, and this model is write model bank;
(4). the model in issue, the application model storehouse.
2. the method for analyzing product association of data service in mobile communication industry according to claim 1 is characterized in that, described step (1) further comprises:
(11). according to the mobile communication business demand, the professional theme that specified data is excavated;
(12). select professional theme corresponding service detail list and related data dimension table;
(13). associated services detail list and related data dimension table;
(14). the data after the association are carried out pre-service, obtain the business datum fairground.
3. the method for analyzing product association of data service in mobile communication industry according to claim 1 is characterized in that, described step (3) further comprises:
(31). according to the assessment inspection parameter of each correlation rule model, select optimum correlation rule model;
(32). invoke script writes model bank in real time with optimum correlation rule model.
4. the method for analyzing product association of data service in mobile communication industry according to claim 1 is characterized in that, described mobile communication business comprises cell phone television services, mobile Ring Back Tone service, mobile ring service, moves complete bent business.
5. the method for analyzing product association of data service in mobile communication industry according to claim 2 is characterized in that, described pre-service comprises:
Business datum in the related data dimension table is cleaned, filtered heavily;
With the format conversion of cleaning, filtering the business datum after weighing is the data layout that the correlation rule model is suitable for.
6. the method for analyzing product association of data service in mobile communication industry according to claim 2 is characterized in that, described Data Mart comprises pretreated day business datum, month business datum and all business datums.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200810039885A CN101620597A (en) | 2008-06-30 | 2008-06-30 | Method for analyzing product association of data service in mobile communication industry |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200810039885A CN101620597A (en) | 2008-06-30 | 2008-06-30 | Method for analyzing product association of data service in mobile communication industry |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101620597A true CN101620597A (en) | 2010-01-06 |
Family
ID=41513839
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200810039885A Pending CN101620597A (en) | 2008-06-30 | 2008-06-30 | Method for analyzing product association of data service in mobile communication industry |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101620597A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521706A (en) * | 2011-12-16 | 2012-06-27 | 北京斯泰威网络科技有限公司 | KPI data analysis method and device for the same |
| CN103150696A (en) * | 2011-12-06 | 2013-06-12 | 中兴通讯股份有限公司 | Method and device for selecting potential customer of target value-added service |
| CN105574087A (en) * | 2015-12-10 | 2016-05-11 | 天津海量信息技术有限公司 | Necessary condition analysis method according to data association rules |
| CN107704929A (en) * | 2016-08-08 | 2018-02-16 | 华为技术有限公司 | A kind of model update method and device |
-
2008
- 2008-06-30 CN CN200810039885A patent/CN101620597A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103150696A (en) * | 2011-12-06 | 2013-06-12 | 中兴通讯股份有限公司 | Method and device for selecting potential customer of target value-added service |
| CN102521706A (en) * | 2011-12-16 | 2012-06-27 | 北京斯泰威网络科技有限公司 | KPI data analysis method and device for the same |
| CN105574087A (en) * | 2015-12-10 | 2016-05-11 | 天津海量信息技术有限公司 | Necessary condition analysis method according to data association rules |
| CN105574087B (en) * | 2015-12-10 | 2018-12-07 | 天津海量信息技术股份有限公司 | Necessary condition analysis method according to data correlation rule |
| CN107704929A (en) * | 2016-08-08 | 2018-02-16 | 华为技术有限公司 | A kind of model update method and device |
| CN107704929B (en) * | 2016-08-08 | 2020-10-23 | 华为技术有限公司 | Model updating method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Adeleye et al. | Evaluation of ICT development and economic growth in Africa | |
| Dietzenbacher et al. | Input–output analysis: the next 25 years | |
| CN103150696A (en) | Method and device for selecting potential customer of target value-added service | |
| Nasir et al. | Information and communication technology-enabled modern services export performances of Asian economies | |
| CN108681970A (en) | Finance product method for pushing, system and computer storage media based on big data | |
| CN107437124A (en) | A kind of operator based on big data analysis complains and trouble correlation analytic method | |
| CN108416701A (en) | A kind of processing method of the Engineering Project Construction real data based on BIM | |
| CN106021239B (en) | A kind of translation quality real-time estimating method | |
| CN103631835A (en) | Interest point information map presenting system and method thereof | |
| CN106803799B (en) | Performance test method and device | |
| CN103198086A (en) | Information processing device, information processing method, and program | |
| CN101103349A (en) | Method for extracting content, content extraction server based on RSS and apparatus for managing the same and system for providing standby screen of mobile communication terminal using the same | |
| CN111008253A (en) | Data model generation method, data warehouse generation device and electronic equipment | |
| CN109598534A (en) | Information determines method and device, electronic equipment and storage medium | |
| CN111191133B (en) | Service search processing method, device and equipment | |
| Masip-Bruin et al. | Unlocking the value of open data with a process-based information platform | |
| CN107045540A (en) | Data processing method and device, server and data handling system | |
| CN110928903B (en) | Data extraction method and device, equipment and storage medium | |
| Pulido-Fernandez et al. | Tourism as a tool for economic development in poor countries | |
| CN109670020A (en) | A kind of voice interactive method, system and device | |
| US20230244862A1 (en) | Form processing method and apparatus, device, and storage medium | |
| CN101620597A (en) | Method for analyzing product association of data service in mobile communication industry | |
| CN112785089A (en) | Agent service configuration method and device, electronic equipment and storage medium | |
| CN116012060A (en) | Customer affinity calculation method, device, equipment and readable storage medium | |
| Carletto et al. | Positioning household surveys for the next decade |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20100106 |