CN109615426A - A kind of marketing method based on Customer clustering, system - Google Patents
A kind of marketing method based on Customer clustering, system Download PDFInfo
- Publication number
- CN109615426A CN109615426A CN201811478602.2A CN201811478602A CN109615426A CN 109615426 A CN109615426 A CN 109615426A CN 201811478602 A CN201811478602 A CN 201811478602A CN 109615426 A CN109615426 A CN 109615426A
- Authority
- CN
- China
- Prior art keywords
- cluster
- sample
- client
- classification
- marketing
- 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
- 238000000034 method Methods 0.000 title claims abstract description 40
- 230000003542 behavioural effect Effects 0.000 claims description 4
- 238000010586 diagram Methods 0.000 description 5
- 230000011218 segmentation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/16—Real estate
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Marketing (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Tourism & Hospitality (AREA)
- Entrepreneurship & Innovation (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Human Resources & Organizations (AREA)
- Probability & Statistics with Applications (AREA)
- Health & Medical Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides a kind of marketing method based on Customer clustering, system, by carrying out classification processing using access duration, access times of the client to platform, it is divided into the client with similar characteristic in same classification cluster, server can be to the client in different classifications cluster, corresponding marketing methods are taken, marketing can be made more targeted, improve marketing efficiency, the work load of sales force is reduced, more targeted marketing methods can also be converted into the promotion of house probability of transaction.
Description
Technical field
The present invention relates to real estate domain more particularly to a kind of marketing methods based on Customer clustering, system.
Background technique
Currently, continuing intimately to develop with real estate domain, number of users is more and more huger on platform, using traditional
Marketing follow-up is individually carried out to each client, can no longer meet actual use demand, also seriously increases the work of sales force
It bears.Therefore, how it is more targeted provide marketing service for different types of client, and then be converted into mentioning for probability of transaction
It rises, is current urgent problem to be solved.
Summary of the invention
A kind of marketing method based on Customer clustering provided by the invention, system, mainly solving the technical problems that: how
It is more targeted to provide marketing service for different types of client.
In order to solve the above technical problems, the present invention provides a kind of marketing method based on Customer clustering, comprising:
S1: when receiving sort instructions, the historical behavior data of each client are obtained from database；The sort instructions
In carry classification cluster number k (k be more than or equal to 2)；The historical behavior data include access duration to platform, access time
Number；
S2: to each client (quantity N, and the N is greater than historical behavior data k) and carries out z-scores mark
Standardization, by the historical behavior data of the client after being standardized, as sample data to be sorted；
S3: from sample data to be sorted, a sample is randomly choosed as first initial cluster center, from remaining N-
In 1 sample, select one with the sample of first initial cluster center lie farthest away as second initial cluster center, from
In remaining N-2 sample, the sample of one with first initial cluster center, second initial cluster center lie farthest away are selected
As third initial cluster center ... ... until selecting k-th of sample as k-th of initial cluster center；
S4: the distance between each sample and each cluster centre in addition to as cluster centre are calculated separately, and will be described each
Sample is divided into itself in immediate classification cluster；
S5: the new cluster centre of each classification cluster is determined according to preset rules；
S6: repeating step S4-S5, until cluster centre is not changing or the number of iterations reaches given threshold
When, export cluster result；
S7: according to cluster result, corresponding marketing methods are selected for the client of different clusters.
Further, the preset rules include:
For each classification cluster, calculate the distance between each sample and remaining sample of the classification cluster, selection and its
The new cluster centre of the remaining sample apart from immediate sample as the classification cluster.
Further, the marketing methods include: supplying system message, control seat system call-on back by phone, notice sale
At least one of personnel's follow-up return visit.
The present invention also provides a kind of marketing systems based on Customer clustering, including user terminal, server, database；
The user terminal is used to acquire the historical behavior data of client, and uploads to the server；
The server is in the historical behavior data for receiving the user terminal uploads, by the historical behavior data
It is stored in the database；
The server executes following steps: S1: when receiving sort instructions, obtaining each client from the database
Historical behavior data；Classification cluster number k is carried in the sort instructions (k is more than or equal to 2)；The historical behavior data packet
Include access duration, the access times to platform；S2: to each client (quantity N, and the N is greater than history row k)
Z-scores standardization is carried out for data, by the historical behavior data of the client after being standardized, as sample to be sorted
Notebook data；S3: from the sample data to be sorted, randomly choosing a sample as first initial cluster center, from surplus
In remaining N-1 sample, select the sample of one and first initial cluster center lie farthest away initial poly- as second
Class center selects one and first initial cluster center, second initial clustering from remaining N-2 sample
The sample of center lie farthest away is initial as k-th up to selecting k-th of sample as third initial cluster center ... ...
Cluster centre；S4: the distance between each sample and each cluster centre in addition to as cluster centre are calculated separately, and will be described each
Sample is divided into itself in immediate classification cluster；S5: the new cluster of each classification cluster is determined according to preset rules
Center；S6: repeating step S4-S5, until cluster centre is not when changing or the number of iterations reaches given threshold, it is defeated
Cluster result out；S7: according to cluster result, corresponding marketing methods are selected for the client of different clusters.
Further, the preset rules include:
For each classification cluster, calculate the distance between each sample and remaining sample of the classification cluster, selection and its
The new cluster centre of the remaining sample apart from immediate sample as the classification cluster.
Further, the marketing system based on Customer clustering further includes seat system；
The server is used to select corresponding marketing methods for the client of different clusters to include: when for first object point
When class cluster, the server is used to control the seat system and carries out phone to all clients in first object classification cluster
It pays a return visit.
Further, the marketing system based on Customer clustering further includes point-of-sale terminal；
The server is used to select corresponding marketing methods for the client of different clusters further include: when for the second target
When classification cluster, the server is used to send the connection of the target customer in the second target classification cluster to the point-of-sale terminal
Mode and classification declaration are paid a return visit with carrying out phone follow-up to the target customer for the point-of-sale terminal.
Further, the server is used to select corresponding marketing methods for the client of different clusters further include: works as needle
When to third target classification cluster, all pushes customer systems of the server for into the third target classification cluster disappear
Breath.
The beneficial effects of the present invention are:
A kind of marketing method based on Customer clustering for there is provided according to the present invention, system, by using client to platform
Duration, access times progress classification processing are accessed, is divided into the client with similar characteristic in same classification cluster, server can
To take corresponding marketing methods to the client in different classifications cluster, marketing can be made more targeted, improve marketing effect
Rate, reduces the work load of sales force, and more targeted marketing methods can also be converted into the promotion of house probability of transaction.
Detailed description of the invention
Fig. 1 is a kind of marketing method flow diagram based on Customer clustering of the embodiment of the present invention one；
Fig. 2 is a kind of communication system architecture schematic diagram based on Customer clustering of the embodiment of the present invention two；
Fig. 3 is another communication system architecture schematic diagram based on Customer clustering of the embodiment of the present invention two.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, below by specific embodiment knot
Closing attached drawing, invention is further described in detail.It should be appreciated that specific embodiment described herein is only used to explain this
Invention, is not intended to limit the present invention.
Embodiment one:
Referring to Figure 1, Fig. 1 is a kind of marketing method based on Customer clustering provided in this embodiment, mainly includes as follows
Step:
S1: when receiving sort instructions, the historical behavior data of each client are obtained from database；It is taken in sort instructions
With classification cluster number k (k is more than or equal to 2)；Historical behavior data include access duration, the access times to platform.
Wherein, sort instructions, which can be, has a periodic progress based on setting time, for example, at interval of three days to data into
One subseries of row, or a subseries is carried out to data at interval of one week.Setting time can be carried out according to practical marketing demand
Setting, such as in order to more frequently being carry out promotion, can set more shorter for setting time, such as be set as one
It.
It should be understood that sort instructions are not limited to aforesaid way, can also be according to the actual situation, sporadically into
Row.
S2: to each client (quantity N, and N is greater than historical behavior data k) and carries out z-scores standardization, will carry out
The historical behavior data of client after standardization, as sample data to be sorted.
In order to better understand, it may refer to customer historical behavioral data as shown in table 1 below:
Table 1
| Client id | Access duration | Access times |
| Sdd122 | 151515 | 11552 |
| Jgh566 | 152265 | 15362 |
| Dgg667 | 292260 | 55326 |
| …… | …… | …… |
Z-scores standardization is carried out to customer historical behavioral data, primarily to eliminating the shadow that dimension analyzes data
It rings, wherein standardized mode can also use existing any other mode, herein with no restrictions.
S3: from sample data to be sorted, a sample is randomly choosed as first initial cluster center, from remaining N-
In 1 sample, select one with the sample of first initial cluster center lie farthest away as second initial cluster center, from
In remaining N-2 sample, the sample of one with first initial cluster center, second initial cluster center lie farthest away are selected
As third initial cluster center ... ... until selecting k-th of sample as k-th of initial cluster center.
For example, sample to be sorted has 10, classification cluster is set as 3, then a sample can be randomly selected as the
One initial cluster center calculates at a distance from first initial cluster center of each sample and this from remaining 9 samples, selects
A sample farthest with distance between first initial cluster center is selected as second initial cluster center, is continued from surplus
Remaining in 8 samples, calculate the distance between each sample and first, second initial cluster center, selection with this first
It is a, between second initial cluster center a farthest sample of distance as third initial cluster center.It can make in this way
The classifying quality obtained finally can more tend to global optimization, avoid currently directly leading by randomly selecting initial cluster center
It causes, the problem of can only achieve local optimum, be unable to reach global optimum, so that classifying quality is more accurate.
Optionally, the mode for calculating distance between sample is as follows:
Assuming that the sample in data set X, X has d attribute or feature, (this example includes access time, access times two
A feature), then data sample xi=(xi1,xi2,…,xid),xj=(xj1,xj2,…xjd), it is sample x respectivelyiAnd xjCorresponding d
The specific value of feature.
Sample xiAnd xjBetween similarity using they the distance between d (xi,xj) indicate, apart from smaller, sample xi
And xjIt is more similar；Bigger, the sample x of distanceiAnd xjSimilitude is smaller.Here, selecting Euclidean distance to characterize to it.
Euclidean distance formula is as follows:
It should be noted that the selection for third initial cluster center, then what is calculated is it at the beginning of with first, second
The sum of the distance between beginning cluster centre.
S4: calculating separately the distance between each sample and each cluster centre in addition to as cluster centre, and by each sample
It is divided into itself in immediate classification cluster.
Continuation is illustrated with above-mentioned example, determined first, second, after third initial cluster center, count
Other 7 samples other than these three cluster centres are calculated, respectively at the distance between these three initial cluster centers, if a certain
The distance between sample and first initial cluster center relative to second, third initial cluster center it is nearest, then should
Sample is divided into the classification cluster where first initial cluster center；If between a certain sample and second initial cluster center
Distance relative to first, third initial cluster center it is nearest, then the sample is divided into second initial clustering
Classification cluster where the heart.
S5: the new cluster centre of each classification cluster is determined according to preset rules.
It after the completion of first time iteration, usually also needs to carry out successive ignition, until the position of cluster centre no longer occurs
Variation, or until the number of iterations of setting reaches.So it needs to be determined that new cluster centre, new cluster centre is according to pre-
If rule is determined, comprising:
For each classification cluster, calculate the distance between each sample and remaining sample of the classification cluster, selection and its
The new cluster centre of the remaining sample apart from immediate sample as the classification cluster.
For example, form three classification clusters after the completion of current iteration, wherein first classification cluster include sample 1, sample 2,
Sample 3, second classification cluster include sample 4, sample 5, sample 6 and sample 7, and third classification cluster includes sample 8, sample 9
And sample 10；It is so directed to first classification cluster, distance d1 of the sample 1 relative to sample 2 and sample 3 can be calculated,
Sample 2 the distance d2 referring now to sample 1 and sample 3, distance d3 of the sample 3 relative to sample 1 and sample 2, if d1 < d2 < d3,
It then can be by sample 1 as new cluster centre.It that is to say and select with remaining sample apart from immediate sample as the classification
The new cluster centre of cluster.Other classification clusters can equally be handled with appellate mode.
S6: repeating step S4-S5, until cluster centre is not changing or the number of iterations reaches given threshold
When, export cluster result.
In order to be optimal classification results, need to carry out successive ignition, until cluster centre is no longer changed, such as
Last cluster centre is that sample 1 recalculates the classification cluster after current iteration, still obtains the poly- of the classification cluster
Class center is sample 1, i.e. cluster centre is no longer changed.
The number of iterations can with practical situations flexible setting, such as be set as 5 times, 10 times, it is 20 inferior.
S7: according to cluster result, corresponding marketing methods are selected for the client of different clusters.
Classification processing is carried out by accessing duration, access times etc. to client, the client of similar access behavior will be divided
It is divided into same classification cluster into same classification cluster, such as by access duration, all more client of access times, when by accessing
Client long, that access times are placed in the middle is divided into another classification cluster, will access the fewer client of duration, access times be divided into and
One classification cluster；It, can be by notifying sales force for each client in all more classification cluster of access duration, access times
Carry out phone follow-up pay a return visit mode, the affection exchange with client can be promoted, can more comprehensively understand client dynamic with
And intention, conveniently facilitate client trading；The client placed in the middle for access duration, access times, can be carried out by seat system
Call-on back by phone reduces the work load of sales force, keeps the telephonic communication of sales force more targeted, passes through seat system pair
Client accesses, and further appreciates that customer action dynamic；The client fewer for access duration, access times, can be with
By server to its user terminal supplying system message, this kind of client is improved to the degree of understanding of related system message, is promoted
Client is inclined to the close friend of this system.
Marketing method provided in an embodiment of the present invention based on client segmentation, when by using client to the access of platform
Long, access times carry out classification processing, are divided into the client with similar characteristic in same classification cluster, server can be to not
With the client in classification cluster, corresponding marketing methods are taken, marketing can be made more targeted, improve marketing efficiency, are reduced
The work load of sales force, more targeted marketing methods can also be converted into the promotion of house probability of transaction.
Embodiment two:
The present embodiment on the basis of example 1, provides a kind of marketing system based on Customer clustering, refers to Fig. 2,
Fig. 2 be a kind of communication system architecture schematic diagram based on Customer clustering provided in this embodiment, the system include user terminal 21,
Server 22, database 23；
Wherein, user terminal 21 is used to acquire the historical behavior data of client, and uploads onto the server 22；
Historical behavior data are stored in number when receiving the historical behavior data of the upload of user terminal 21 by server 22
According in library 23；
Server 22 executes following steps: when receiving sort instructions, going through for each client S1: is obtained from database 23
History behavioral data；Classification cluster number k is carried in sort instructions (k is more than or equal to 2)；Historical behavior data include the visit to platform
Ask duration, access times；S2: to each client (quantity N, and N is greater than historical behavior data k) and carries out z-scores standard
Change, by the historical behavior data of the client after being standardized, as sample data to be sorted；S3: from sample to be sorted
In data, a sample is randomly choosed as first initial cluster center, from remaining N-1 sample, select one and the
The sample of one initial cluster center lie farthest away is as second initial cluster center, from remaining N-2 sample, selects one
It is a with first initial cluster center, the sample of second initial cluster center lie farthest away is as in third initial clustering
The heart ... ... is until select k-th of sample as k-th of initial cluster center；S4: it calculates separately in addition to as cluster centre
The distance between each sample and each cluster centre, and each sample is divided into itself in immediate classification cluster；S5: it presses
The new cluster centre of each classification cluster is determined according to preset rules；S6: repeating step S4-S5, until cluster centre does not exist
It changes or when the number of iterations reaches given threshold, exports cluster result；S7: according to cluster result, for the visitor of different clusters
Family selects corresponding marketing methods.
Wherein preset rules include: calculated for each classification cluster the classification cluster each sample and remaining sample it
Between distance, select with new cluster centre of remaining sample apart from immediate sample as the classification cluster.
Optionally, Fig. 3 is referred to, Fig. 3 is another communication system architecture based on Customer clustering provided in this embodiment
Schematic diagram further includes seat system 24；
Server 22 is used to select corresponding marketing methods for the client of different clusters to include: to classify when for first object
When cluster, server 22 is used to control seat system 24 and carries out call-on back by phone to all clients in first object classification cluster.
With continued reference to Fig. 3, being somebody's turn to do the marketing system based on client segmentation further includes point-of-sale terminal 25；
Server 22 is used to select corresponding marketing methods for the client of different clusters further include: when for the second target point
When class cluster, server 22 is used to send the contact method of the target customer in the second target classification cluster to point-of-sale terminal 25 and divides
Class declaration is paid a return visit with carrying out phone follow-up to target customer for point-of-sale terminal 25.
Wherein, the contact method of client can be obtained based on client's data of filing, and usually can store in database 23
In, server 22 can be acquired from database.Point-of-sale terminal 25 usually can be house agent, third party sells people
Member etc. uses.
When being directed to third target classification cluster, server 22 is for all pushes customer systems into third target classification cluster
System message.
Such as access duration, the more client of access times is divided into the second target classification cluster, will access duration,
Access times client placed in the middle is divided into first object classification cluster, will access the fewer client of duration, access times and is divided into
Third target classification cluster；For each client in access duration, all more second target classification cluster of access times, Ke Yitong
It crosses and sales force is notified to carry out the mode that phone follow-up is paid a return visit, the affection exchange with client can be promoted, it can be more comprehensive
Understand client's dynamic and intention, conveniently facilitates client trading；Second target classification placed in the middle for access duration, access times
Each client in cluster can carry out call-on back by phone by seat system, reduce the work load of sales force, make sales force's
Telephonic communication is more targeted, is accessed by seat system to client, further appreciates that customer action dynamic；For access
Each client in the fewer third target classification cluster of duration, access times can be pushed away by server to its user terminal
System message is sent, this kind of client is improved to the degree of understanding of related system message, promotes client and the close friend of this system is inclined to.
Obviously, those skilled in the art should be understood that each module of aforementioned present invention or each step can be with general
Computing device realizes that they can be concentrated on a single computing device, or be distributed in constituted by multiple computing devices
On network, optionally, they can be realized with the program code that computing device can perform, it is thus possible to be stored in
It is performed by computing device in computer storage medium (ROM/RAM, magnetic disk, CD), and in some cases, it can be with not
The sequence being same as herein executes shown or described step, or they are fabricated to each integrated circuit modules, or
Person makes multiple modules or steps in them to single integrated circuit module to realize.So the present invention is not limited to appoint
What specific hardware and software combines.
The above content is specific embodiment is combined, further detailed description of the invention, and it cannot be said that this hair
Bright specific implementation is only limited to these instructions.For those of ordinary skill in the art to which the present invention belongs, it is not taking off
Under the premise of from present inventive concept, a number of simple deductions or replacements can also be made, all shall be regarded as belonging to protection of the invention
Range.
Claims (8)
1. a kind of marketing method based on Customer clustering, which is characterized in that the marketing method based on Customer clustering includes:
S1: when receiving sort instructions, the historical behavior data of each client are obtained from database；It is taken in the sort instructions
With classification cluster number k (k is more than or equal to 2)；The historical behavior data include access duration, the access times to platform；
S2: z-scores standardization is carried out to the historical behavior data of each client, by the client after being standardized
Historical behavior data, as sample data to be sorted；The quantity of each client is N, and the N is greater than the k；
S3: from the sample data to be sorted, a sample is randomly choosed as first initial cluster center, from remaining N-
In 1 sample, select one with the sample of first initial cluster center lie farthest away as in second initial clustering
The heart selects one and first initial cluster center, second initial cluster center from remaining N-2 sample
The sample of lie farthest away is as third initial cluster center ... ... until selecting k-th of sample as k-th of initial clustering
Center；
S4: calculating separately the distance between each sample and each cluster centre in addition to as cluster centre, and by each sample
It is divided into itself in immediate classification cluster；
S5: the new cluster centre of each classification cluster is determined according to preset rules；
S6: repeating step S4-S5, until cluster centre is not when changing or the number of iterations reaches given threshold, it is defeated
Cluster result out；
S7: according to cluster result, corresponding marketing methods are selected for the client of different clusters.
2. as described in claim 1 based on the marketing method of Customer clustering, which is characterized in that the preset rules include:
For each classification cluster, the distance between each sample and remaining sample of the classification cluster, selection and remaining sample are calculated
This described new cluster centre apart from immediate sample as the classification cluster.
3. as claimed in claim 1 or 2 based on the marketing method of Customer clustering, which is characterized in that the marketing methods include:
Supplying system message, control seat system call-on back by phone notify that sales force's follow-up is paid a return visit.
4. a kind of marketing system based on Customer clustering, which is characterized in that the marketing system based on Customer clustering includes using
Family terminal, server, database；
The user terminal is used to acquire the historical behavior data of client, and uploads to the server；
The server stores the historical behavior data in the historical behavior data for receiving the user terminal uploads
In the database；
The server executes following steps: S1: when receiving sort instructions, obtaining going through for each client from the database
History behavioral data；Classification cluster number k is carried in the sort instructions (k is more than or equal to 2)；The historical behavior data include pair
Access duration, the access times of platform；S2: to each client (quantity N, and the N is greater than historical behavior number k)
According to z-scores standardization is carried out, by the historical behavior data of the client after being standardized, as sample number to be sorted
According to；S3: from the sample data to be sorted, a sample is randomly choosed as first initial cluster center, from remaining N-1
In a sample, select one with the sample of first initial cluster center lie farthest away as in second initial clustering
The heart selects one and first initial cluster center, second initial cluster center from remaining N-2 sample
The sample of lie farthest away is as third initial cluster center ... ... until selecting k-th of sample as k-th of initial clustering
Center；S4: calculating separately the distance between each sample and each cluster centre in addition to as cluster centre, and by each sample
It is divided into itself in immediate classification cluster；S5: the new cluster centre of each classification cluster is determined according to preset rules；
S6: repeating step S4-S5, until cluster centre, not when changing or the number of iterations reaches given threshold, output is poly-
Class result；S7: according to cluster result, corresponding marketing methods are selected for the client of different clusters.
5. as claimed in claim 4 based on the marketing system of Customer clustering, which is characterized in that the preset rules include:
For each classification cluster, the distance between each sample and remaining sample of the classification cluster, selection and remaining sample are calculated
This described new cluster centre apart from immediate sample as the classification cluster.
6. as described in claim 4 or 5 based on the marketing system of Customer clustering, which is characterized in that described to be based on Customer clustering
Marketing system further include seat system；
It includes: when for first object classification cluster that the server, which selects corresponding marketing methods for the client for different clusters,
When, the server is used to control the seat system and carries out phone time to all clients in first object classification cluster
It visits.
7. as claimed in claim 6 based on the marketing system of Customer clustering, which is characterized in that the battalion based on Customer clustering
Pin system further includes point-of-sale terminal；
The server is used to select corresponding marketing methods for the client of different clusters further include: when for the second target classification
When cluster, the server is used to send the contact method of the target customer in the second target classification cluster to the point-of-sale terminal
And classification declaration, it is paid a return visit with carrying out phone follow-up to the target customer for the point-of-sale terminal.
8. as claimed in claim 6 based on the marketing system of Customer clustering, which is characterized in that the server is used for for not
Client with cluster selects corresponding marketing methods further include: when being directed to third target classification cluster, the server is used for institute
State all pushes customer system messages in third target classification cluster.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811478602.2A CN109615426A (en) | 2018-12-05 | 2018-12-05 | A kind of marketing method based on Customer clustering, system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811478602.2A CN109615426A (en) | 2018-12-05 | 2018-12-05 | A kind of marketing method based on Customer clustering, system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109615426A true CN109615426A (en) | 2019-04-12 |
Family
ID=66007072
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811478602.2A Pending CN109615426A (en) | 2018-12-05 | 2018-12-05 | A kind of marketing method based on Customer clustering, system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109615426A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110070297A (en) * | 2019-04-28 | 2019-07-30 | 重庆锐云科技有限公司 | Tourist classification marketing management method, Scenery Management System, server |
| CN110113499A (en) * | 2019-05-10 | 2019-08-09 | 南京硅基智能科技有限公司 | Virtual mobile phone intelligent call method and system based on cloud |
| CN110111158A (en) * | 2019-05-16 | 2019-08-09 | 创络(上海)数据科技有限公司 | The Marketing Design method, apparatus and storage medium of life cycle or Development phase |
| CN111563628A (en) * | 2020-05-09 | 2020-08-21 | 重庆锐云科技有限公司 | Real estate customer transaction time prediction method, device and storage medium |
| CN115953166A (en) * | 2022-12-27 | 2023-04-11 | 鑫恒绅企业服务(无锡)有限公司 | Customer information management method and system based on big data intelligent matching |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8782045B1 (en) * | 2010-04-15 | 2014-07-15 | Google Inc. | Evaluating techniques for clustering geographic entities |
| CN104376057A (en) * | 2014-11-06 | 2015-02-25 | 南京邮电大学 | Self-adaptation clustering method based on maximum distance, minimum distance and K-means |
| CN105389713A (en) * | 2015-10-15 | 2016-03-09 | 南京大学 | Mobile data traffic package recommendation algorithm based on user historical data |
| CN106127519A (en) * | 2016-06-24 | 2016-11-16 | 武汉斗鱼网络科技有限公司 | A kind of live platform user divided method based on K Means algorithm and system |
| CN107305637A (en) * | 2016-04-21 | 2017-10-31 | 华为技术有限公司 | Data clustering method and device based on K-Means algorithms |
| CN107423636A (en) * | 2017-07-06 | 2017-12-01 | 北京航空航天大学 | A kind of difference privacy K mean cluster method based on MapReduce |
| US20180108048A1 (en) * | 2016-10-19 | 2018-04-19 | Samsung Sds Co., Ltd. | Method, apparatus and system for recommending contents |
-
2018
- 2018-12-05 CN CN201811478602.2A patent/CN109615426A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8782045B1 (en) * | 2010-04-15 | 2014-07-15 | Google Inc. | Evaluating techniques for clustering geographic entities |
| CN104376057A (en) * | 2014-11-06 | 2015-02-25 | 南京邮电大学 | Self-adaptation clustering method based on maximum distance, minimum distance and K-means |
| CN105389713A (en) * | 2015-10-15 | 2016-03-09 | 南京大学 | Mobile data traffic package recommendation algorithm based on user historical data |
| CN107305637A (en) * | 2016-04-21 | 2017-10-31 | 华为技术有限公司 | Data clustering method and device based on K-Means algorithms |
| CN106127519A (en) * | 2016-06-24 | 2016-11-16 | 武汉斗鱼网络科技有限公司 | A kind of live platform user divided method based on K Means algorithm and system |
| US20180108048A1 (en) * | 2016-10-19 | 2018-04-19 | Samsung Sds Co., Ltd. | Method, apparatus and system for recommending contents |
| CN107423636A (en) * | 2017-07-06 | 2017-12-01 | 北京航空航天大学 | A kind of difference privacy K mean cluster method based on MapReduce |
Non-Patent Citations (1)
| Title |
|---|
| 舒宁: "《模式识别的理论与方法》", 31 December 2004 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110070297A (en) * | 2019-04-28 | 2019-07-30 | 重庆锐云科技有限公司 | Tourist classification marketing management method, Scenery Management System, server |
| CN110113499A (en) * | 2019-05-10 | 2019-08-09 | 南京硅基智能科技有限公司 | Virtual mobile phone intelligent call method and system based on cloud |
| CN110111158A (en) * | 2019-05-16 | 2019-08-09 | 创络(上海)数据科技有限公司 | The Marketing Design method, apparatus and storage medium of life cycle or Development phase |
| CN111563628A (en) * | 2020-05-09 | 2020-08-21 | 重庆锐云科技有限公司 | Real estate customer transaction time prediction method, device and storage medium |
| CN115953166A (en) * | 2022-12-27 | 2023-04-11 | 鑫恒绅企业服务(无锡)有限公司 | Customer information management method and system based on big data intelligent matching |
| CN115953166B (en) * | 2022-12-27 | 2024-04-02 | 鑫恒绅企业服务(无锡)有限公司 | Customer information management method and system based on big data intelligent matching |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109615426A (en) | A kind of marketing method based on Customer clustering, system | |
| CN109918454B (en) | Method and device for embedding nodes into relational network graph | |
| WO2021174944A1 (en) | Message push method based on target activity, and related device | |
| CN110046929B (en) | Fraudulent party identification method and device, readable storage medium and terminal equipment | |
| CN108710634B (en) | Protocol file pushing method and terminal equipment | |
| WO2019196255A1 (en) | Time period selection method and device for telephone sales, terminal equipment and medium | |
| CN106228386A (en) | A kind of information-pushing method and device | |
| CN111210341A (en) | Method and device for determining service quota | |
| JP6288662B1 (en) | Performance prediction management system and method | |
| WO2019085704A1 (en) | Method and apparatus for increasing the number of active users | |
| CN111311310B (en) | Advertisement order pushing method and device, storage medium and electronic device | |
| CN113743968A (en) | Information delivery method, device and equipment | |
| CN105138897A (en) | Method and device for determining user right | |
| CN112416590A (en) | Server system resource adjusting method and device, computer equipment and storage medium | |
| CN112085332A (en) | Data processing method, data processing device, computer equipment and storage medium | |
| CN108629582A (en) | Method for processing business and device | |
| CN110210884A (en) | Determine the method, apparatus, computer equipment and storage medium of user characteristic data | |
| Fry | Ebook rate of use in OhioLINK: A ten-year study of local and consortial use of publisher packages in Ohio | |
| CN111753208B (en) | Method, device, medium and electronic equipment for determining convergence of comparable attributes of users | |
| CN105653645B (en) | Network information attention degree evaluation method and device | |
| CN114298825A (en) | Method and device for extremely evaluating repayment volume | |
| CN112818235A (en) | Violation user identification method and device based on associated features and computer equipment | |
| CN111026816A (en) | High-net-value customer group identification method and device based on knowledge graph and storage medium | |
| CN111681090A (en) | Account grouping method and device of business system, terminal equipment and storage medium | |
| CN113934920A (en) | Target information pushing method and device and storage medium |
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: 20190412 |