CN107590213A - Mixing commending system based on mobile phone big data - Google Patents
Mixing commending system based on mobile phone big data Download PDFInfo
- Publication number
- CN107590213A CN107590213A CN201710755680.1A CN201710755680A CN107590213A CN 107590213 A CN107590213 A CN 107590213A CN 201710755680 A CN201710755680 A CN 201710755680A CN 107590213 A CN107590213 A CN 107590213A
- Authority
- CN
- China
- Prior art keywords
- data
- module
- analysis
- mobile phone
- algorithm
- 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
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of mixing commending system based on mobile phone big data, including data acquisition module, data processing module, data analysis module, information display module and info push module.Regard cellphone subscriber as a circle with social bond, recommended with good friend's interest-degree in circle, and combine the mixing proposed algorithm to be formed using two kinds of proposed algorithms and dig into row operational analysis, to reach the effect preferably recommended.Simultaneously by Hadoop big data analysis platforms, distributed program is developed using it, realizes the function of distributed type high speed processing data, user mobile phone big data is stored and operational analysis.
Description
Technical field
The present invention relates to a kind of mixing commending system based on mobile phone big data, it is mainly used in analysis mining mobile phone big data
User interest information, realize the personalized recommendation to user.
Background technology
In recent years, because the high speed development of information technology, scientific domain generate with commercial field (such as common carrier)
The quite huge data of scale, and these data are in the constantly quick increased stage.In face of the data of magnanimity, big data is deposited
Storage with processing face huge challenge, how fast and effectively to store with processing be present network technology a focus.
Data mining (Data Mining), refer to be hidden in wherein information by algorithm search from substantial amounts of data
Process, by classifying (Classification), clustering (Clustering), association (Association), prediction
(Predicton) and visualization (Visualization) scheduling algorithm and technology find out different users, analysis consumer hobby and
The method of behavior, so as to create huge value for business and market.
Personalized recommendation is the popular domain of current data Research on Mining, and it can be very good to lift Consumer's Experience, so as to
Enterprise is marketed.There are some shortcomings in single personalized recommendation algorithm, the quality and efficiency of recommendation still have improved space,
And traditional personalized recommendation algorithm, relation is to have ignored the relation between user between user is handled.
The content of the invention
The problem of for the above, the present invention proposes a kind of mixing commending system based on mobile phone big data, by cellphone subscriber
Regard a circle with social bond as, recommended with good friend's interest-degree in circle, and use two kinds of proposed algorithm knots
The mixing proposed algorithm collectively formed, make recommendation effect more accurate.There is distribution simultaneously with reference to Hadoop software platforms
The design feature of rowization, distributed program is developed using it, realizes the function of distributed type high speed processing data, it is big to user mobile phone
Data are stored and operational analysis.
Data acquisition module, data processing module, data analysis module, information display module and information is specifically included to push
Module.
The data acquisition module, added by the collection desensitization of mobile communication operator big data co-operation platform interface timing
User mobile phone data after close, then it is transferred to the data processing module.
The data processing module, store and pre-process the data that the data collecting module collected arrives；
The data analysis module, the data that the data processing module is stored and pre-processed are extracted, are first disappeared according to user
Flat cluster all customer groups of water wasting is segmented, and reuses two kinds of proposed algorithms and combines the mixing proposed algorithm to be formed and is transported
Point counting is analysed, and the result of operational analysis is inputted into described information display module and described information pushing module；
Described information display module, receive the operational analysis result of the data analysis module, there is provided customer group integrally counts
Independently inquired about with individual subscriber according to information analysis；
Described information pushing module, the operational analysis result of the data analysis module is received, and according to analysis result master
Trend cellphone subscriber targetedly pushes commodity and service, realizes that supply and demand chain type is predicted, and according to user feedback, continues to optimize and is
System.
The mixing proposed algorithm, it is made up of community discovery algorithm and association algorithm both proposed algorithms；Described in input
All customer groups are clustered to the result of subdivision according to customer consumption level, it is similar first to carry out the community discovery algorithm output interest
Customer group, then different customer groups is excavated according to the association algorithm, output user attribute interested, so
To targetedly recommending user and input described information display module and described information pushing module.
The community discovery algorithm uses GN algorithms, and the association algorithm uses Apriori algorithm.
The data processing module and the data analysis module are by Hadoop big data analysis platforms, are opened using it
Distributed program is sent out, realizes the function of distributed type high speed processing data, user mobile phone big data is stored and operational analysis.
Advantages of the present invention and beneficial effect：
Two kinds of proposed algorithms combine the mixing proposed algorithm to be formed, and have reached the effect that user is grouped by interest first
Fruit, the rule of user interest is secondly found out with association rule algorithm in each group, both, which are combined, greatly improves recommendation
The degree of accuracy, the effect preferably recommended can be reached.
Regard cellphone subscriber as a circle with social bond, recommended with good friend's interest-degree in circle, pushed away
It is more accurate to recommend effect.
Using the powerful data storage capacities of Hadoop platform, the storage of magnanimity level cellphone subscriber's data is realized.And utilize
The powerful data analysis computing capability of Hadoop platform, find that algorithm, association algorithm realize magnanimity level mobile phone using communities of users
The calculating analysis of user data parallelization, significant increase because hardware deficiency cause processing speed excessively slow the problem of.It is final to realize
Personalized commodity and service are pushed to cellphone subscriber, improve the value of mobile phone big data.
Brief description of the drawings
Fig. 1 is the overall pie graph of the embodiment of the present invention；
Fig. 2 is the commending system hierarchical chart of the embodiment of the present invention；
Fig. 3 is the step process of community discovery algorithm；
Fig. 4 is that GN algorithms side betweenness calculates schematic diagram；
Fig. 5 is the step process of association algorithm；
Fig. 6 is based on community discovery and the explanation figure for associating mixing proposed algorithm.
Embodiment
Below in conjunction with the accompanying drawing in inventive embodiments, the technical scheme in the embodiment of the present invention is carried out clear, detailed
Ground describes.Described embodiment is only the part of the embodiment of the present invention.
As shown in figure 1, the mixing commending system based on mobile phone big data include data acquisition module, data processing module,
Data analysis module, information display module and info push module.
Data acquisition module, after the collection desensitization encryption of mobile communication operator big data co-operation platform interface timing
User mobile phone data.The base attribute of user takes 1 expression user base attribute to gather (name, age, sex, occupation, hand
Machine number, in net duration, set meal)；The base attribute of user take the behavior property of 2 expression users gather (internet behavior information,
Call behavioural information, note data, moon flow, consumption data)；The base attribute of user takes the label data of 3 expression users
Gather (terminal preferences, APP preferences, consumption feature, interest, position data).
Data processing module, store the data that simultaneously preprocessed data acquisition module collects.
Data analysis module, the data that extraction data processing module is stored and pre-processed first will according to customer consumption level
All customer group cluster subdivisions, reuse GN algorithms and carry out communities of users discovery, the similar customer group of output interest；Then will not
Same customer group carries out Apriori algorithm association rule mining, output user attribute interested respectively, and then user is carried out
Targetedly recommend.
Data processing module and data analysis module are by Hadoop big data analysis platforms, utilize its exploitation distribution
Program, the function of distributed type high speed processing data is realized, user mobile phone big data is stored and operational analysis.
Information display module, receive the operational analysis result of data analysis module, there is provided customer group overall data information point
Analysis is independently inquired about with individual subscriber.
Info push module, the operational analysis result of data analysis module is received, and according to analysis result actively to mobile phone
User targetedly pushes commodity and service, realizes that supply and demand chain type is predicted, and according to user feedback, continue to optimize system.
Fig. 2 is the hierarchical chart of commending system, including accumulation layer, functional layer and application layer.Accumulation layer includes data
Collection and the storage of data, and the preprocessing function of data；Functional layer includes three algoritic modules：Statistical module, analysis mould
Block and excavation module；Application layer includes two application modules, is information display module and pushing module respectively.
Wherein, the excavation module of accumulation layer includes clustering algorithm and the mixing comprising community discovery algorithm and association algorithm pushes away
Recommend algorithm；Analysis module realizes the monitoring to each item data of system, and can carry out abnormal alarm function；Statistical module is according to data
The various statistical functions of storehouse data, different statistics is realized for different functional requirements.
The pushing module of application layer is based on mobile phone big data and data mining algorithm, realizes actively to cellphone subscriber group
Personalized push, realize that supply and demand chain type is predicted, and according to user feedback, continue to optimize system；Display module is used for user
Group's overall data information analysis is independently inquired about with individual subscriber.
The data storage of accumulation layer, the data taken are stored, the extractions of data, loading and clear are carried out to initial data
Wash, be further ready for analysis；Data acquisition, the data after collection cellphone subscriber's desensitization encryption, its all data are logical
The user mobile phone data crossed after the collection desensitization encryption of mobile phone common carrier big data co-operation platform interface timing.Accumulation layer work(
It can be the preprocessing function for realizing the collections of data, storage and data.
Fig. 3 is community discovery algorithm GN algorithm performs flow chart, when carrying out community's group division, into template degree letter
Number Q,Criterion as community's division.Wherein, if a network is divided into n community, can be formed
One n × n symmetrical matrixes e, eijFor the element in symmetrical matrix, the node that it is represent in community i connects with the node in community j
The ratio when all in accounting for whole network connect；eiiRepresent cornerwise each element；ai=∑jeijWhat is represented is in the matrix
The i-th row element and.
It is as follows that community discovery based on GN algorithms performs step：
S11. in a network, side betweenness N of each edge relative to all possible source node is calculated using Floyd algorithms,
Then the high side of N values is deleted, for the first division in fission process, the template degree Q and figure for calculating primary network are tied
Structure.
S12. for remaining side in network, repeat step S11, until all sides are calculated in network, then, will have
There is final division result of the network structure corresponding to maximum Q values as the network.
Fig. 4 is that GN algorithms side betweenness calculates schematic diagram.Shortest path in search for networks between all nodes, count institute
Have while while betweenness, finally count maximum side betweenness then delete, that is, obtain finally divide result, that is, community divide tie
Fruit.
As shown in figure 5, the step process of association algorithm Apriori algorithm：
S21. mobile phone big data user's matrix D in scan database, calculating the supports of user's all properties, (support is
The probability that family attribute A and user property B occurs simultaneously, Support (A → B)=P (A ∩ B)), and given minimum support threshold value
(foundation of the setting of threshold size is to filter out too small support), if greater than given minimum support threshold value, then
The attribute is defined as frequent episode, it is hereby achieved that all user properties frequent 1- item collections set L1 (if one collection
K element is included in conjunction, then this k item collection meets that the collection of minimum support threshold value is collectively referred to as frequent k item collections).
For S22.Apriori algorithms by the basis of upper frequent (the k-1)-item collection once generated, link generates new time
K- item collections are selected, k element is included in mono- set of k.
S23. when not new frequent item set produces, i.e., when K reaches maximum, algorithm terminates, and obtains the strong of user
Correlation rule.
As shown in fig. 6, communities of users finds the mixing proposed algorithm that algorithm is combined with correlation rule proposed algorithm, number is inputted
According to using GN algorithms to carry out communities of users discovery, the similar customer group of output interest；Then user items matrix is obtained, will not
Same customer group carries out Apriori algorithm association rule mining, exports user's recommendation list.
Claims (4)
1. the mixing commending system based on mobile phone big data, it is characterised in that：Including data acquisition module, data processing module,
Data analysis module, information display module and info push module；
The data acquisition module, after the collection desensitization encryption of mobile communication operator big data co-operation platform interface timing
User mobile phone data, then be transferred to the data processing module；
The data processing module, store and pre-process the data that the data collecting module collected arrives；
The data analysis module, the data that the data processing module is stored and pre-processed are extracted, first according to customer consumption water
Flat cluster all customer groups is segmented, and reuses two kinds of proposed algorithms and combines the mixing proposed algorithm to be formed progress computing point
Analysis, and the result of operational analysis is inputted into described information display module and described information pushing module；
Described information display module, receive the operational analysis result of the data analysis module, there is provided customer group overall data is believed
Breath analysis is independently inquired about with individual subscriber；
Described information pushing module, receives the operational analysis result of the data analysis module, and according to analysis result actively to
Cellphone subscriber targetedly pushes commodity and service, realizes that supply and demand chain type is predicted, and according to user feedback, continue to optimize system.
2. the mixing commending system according to claim 1 based on mobile phone big data, it is characterised in that：The mixing is recommended
Algorithm, it is made up of community discovery algorithm and association algorithm both proposed algorithms；Input it is described according to customer consumption level by institute
The result for thering is user's clustering class to segment, the similar customer group of the community discovery algorithm output interest is first carried out, then will be different
Customer group is excavated according to the association algorithm, output user attribute interested, and then is obtained to user targetedly
Recommend and input described information display module and described information pushing module.
3. the mixing commending system according to claim 2 based on mobile phone big data, it is characterised in that：The community discovery
Algorithm uses GN algorithms, and the association algorithm uses Apriori algorithm.
4. the mixing commending system according to claim 1 based on mobile phone big data, it is characterised in that：The data processing
Module and the data analysis module are by Hadoop big data analysis platforms, develop distributed program using it, realize and divide
The function of cloth process data at high speeds, is stored and operational analysis to user mobile phone big data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710755680.1A CN107590213A (en) | 2017-08-29 | 2017-08-29 | Mixing commending system based on mobile phone big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710755680.1A CN107590213A (en) | 2017-08-29 | 2017-08-29 | Mixing commending system based on mobile phone big data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107590213A true CN107590213A (en) | 2018-01-16 |
Family
ID=61050273
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710755680.1A Pending CN107590213A (en) | 2017-08-29 | 2017-08-29 | Mixing commending system based on mobile phone big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107590213A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108650166A (en) * | 2018-03-19 | 2018-10-12 | 安徽锐欧赛智能科技有限公司 | A kind of timely communication data analysis and message push management system |
| CN110289049A (en) * | 2019-06-19 | 2019-09-27 | 江南大学 | A method of analysis communication path and the heat-resisting sexual intercourse of lipase |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130204948A1 (en) * | 2012-02-07 | 2013-08-08 | Cloudera, Inc. | Centralized configuration and monitoring of a distributed computing cluster |
| CN103714139A (en) * | 2013-12-20 | 2014-04-09 | 华南理工大学 | Parallel data mining method for identifying a mass of mobile client bases |
| CN105913342A (en) * | 2016-04-08 | 2016-08-31 | 上海旭薇物联网科技有限公司 | Smart community system based on big data mining algorithm |
| CN105956048A (en) * | 2016-04-27 | 2016-09-21 | 上海遥薇(集团)有限公司 | Community service big data algorithm mining system |
| CN106296305A (en) * | 2016-08-23 | 2017-01-04 | 上海海事大学 | Electric business website real-time recommendation System and method under big data environment |
-
2017
- 2017-08-29 CN CN201710755680.1A patent/CN107590213A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130204948A1 (en) * | 2012-02-07 | 2013-08-08 | Cloudera, Inc. | Centralized configuration and monitoring of a distributed computing cluster |
| CN103714139A (en) * | 2013-12-20 | 2014-04-09 | 华南理工大学 | Parallel data mining method for identifying a mass of mobile client bases |
| CN105913342A (en) * | 2016-04-08 | 2016-08-31 | 上海旭薇物联网科技有限公司 | Smart community system based on big data mining algorithm |
| CN105956048A (en) * | 2016-04-27 | 2016-09-21 | 上海遥薇(集团)有限公司 | Community service big data algorithm mining system |
| CN106296305A (en) * | 2016-08-23 | 2017-01-04 | 上海海事大学 | Electric business website real-time recommendation System and method under big data environment |
Non-Patent Citations (1)
| Title |
|---|
| 栗欢: "基于社区发现和关联规则的论文混合推荐技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108650166A (en) * | 2018-03-19 | 2018-10-12 | 安徽锐欧赛智能科技有限公司 | A kind of timely communication data analysis and message push management system |
| CN110289049A (en) * | 2019-06-19 | 2019-09-27 | 江南大学 | A method of analysis communication path and the heat-resisting sexual intercourse of lipase |
| CN110289049B (en) * | 2019-06-19 | 2021-06-25 | 江南大学 | Method for analyzing heat resistance relation between communication path and lipase |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105512242B (en) | A kind of parallel recommendation method based on social network structure | |
| Rathore et al. | DeepCachNet: A proactive caching framework based on deep learning in cellular networks | |
| CN109345263A (en) | Predict the method and system of customer satisfaction | |
| CN105007171A (en) | User data analysis system and method based on big data in communication field | |
| CN108537134A (en) | A kind of video semanteme scene cut and mask method | |
| CN106778876A (en) | User classification method and system based on mobile subscriber track similitude | |
| Ahmed et al. | Exploring nested ensemble learners using overproduction and choose approach for churn prediction in telecom industry | |
| CN107545315A (en) | Order processing method and device | |
| CN105281925B (en) | The method and apparatus that network service groups of users divides | |
| Xiao et al. | Churn prediction in customer relationship management via GMDH-based multiple classifiers ensemble | |
| CN103138981A (en) | Method and device for social network service analysis | |
| CN104077723A (en) | Social network recommending system and social network recommending method | |
| CN105975479A (en) | Tag library-based telecom user interest degree analysis method and system | |
| CN109558992A (en) | Based on sale peak value prediction technique, device, equipment and the storage medium from the machine of dealer | |
| CN107247762A (en) | A kind of international logistics circuit recommendation method | |
| CN109558991B (en) | Commodity channel quantity recommendation method, device, equipment and storage medium based on vending machine | |
| CN107704868A (en) | Tenant group clustering method based on Mobile solution usage behavior | |
| CN110009416A (en) | A kind of system based on big data cleaning and AI precision marketing | |
| CN107590213A (en) | Mixing commending system based on mobile phone big data | |
| CN108537291A (en) | A kind of sample classification method and apparatus | |
| CN110169019A (en) | The network switch and Database Systems that database function defines | |
| CN106649743A (en) | Method and system for storing and sharing creative idea classified brain library resources | |
| CN110288465A (en) | Object determines method and device, storage medium, electronic device | |
| CN112182031B (en) | Data query method and device, storage medium and electronic device | |
| CN105677888A (en) | Service preference identification method based on user time fragments |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180116 |