CN109919652A - User group's classification method, device, equipment and storage medium - Google Patents
User group's classification method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN109919652A CN109919652A CN201910046571.1A CN201910046571A CN109919652A CN 109919652 A CN109919652 A CN 109919652A CN 201910046571 A CN201910046571 A CN 201910046571A CN 109919652 A CN109919652 A CN 109919652A
- Authority
- CN
- China
- Prior art keywords
- user
- label
- data
- behavior
- target
- 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 46
- 230000006399 behavior Effects 0.000 claims description 90
- 230000015654 memory Effects 0.000 claims description 14
- 238000013507 mapping Methods 0.000 claims description 4
- 238000007405 data analysis Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 9
- 238000004891 communication Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000013523 data management Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 241000282326 Felis catus Species 0.000 description 1
- 206010018265 Gigantism Diseases 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000010845 search algorithm Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to big data analysis, disclosing a kind of user group's classification method, device, equipment and storage medium, this method includes burying the user data and device data of acquisition user by the way that client is preset；Extract equipment identifies and searches whether there is historical user's label associated with device identification from device data；The corresponding target user's label of user is then obtained according to user data if it does not exist；Target user's label is input to default crowd's Definition Model to determine target group's type belonging to user, due to being first to search the corresponding historical user's label of user according to collected device data, continue to carry out tag definition acquisition user tag to user according to user data if not finding, then the label input crowd's Definition Model that will acquire again and crowd's type according to belonging to model output result finally determining user, to realize the accurate positionin to user crowd's type, effective foundation is provided to user's progress pertinent service to be subsequent.
Description
Technical field
The present invention relates to big data analysis technical field more particularly to a kind of user group's classification method, device, equipment and
Storage medium.
Background technique
Current all kinds of electric business websites (such as day cat and Jingdone district), application client (Application, App) have needle
It carries out user to different user to draw a portrait user's targeted promotions of analysis, functions such as personalized recommendation, but in fact since user disappears
Take the gigantism of group, the diversification of service request and consumption demand, user has begun with different demands, different users
Although having different characteristics, age, income, the individual characteies such as consuming capacity and rule of conduct, between same type of user
There are more or less general character.Currently based on the promotion strategy of user's Portrait brand technology from user personality, only it is merely
It has been directed to the single user of a small range, there is no be directed to certain a kind of user group again or intersect the offer of user group more
Therefore how accurate information popularization service carries out crowd to different users or group positions, to be different types of people
Group provides and more targetedly services, with regard to becoming a problem urgently to be resolved at present.
Above content is only used to facilitate the understanding of the technical scheme, and is not represented and is recognized that above content is existing skill
Art.
Summary of the invention
The main purpose of the present invention is to provide a kind of user group's classification method, device, equipment and storage medium, purports
Crowd or group's positioning can not be carried out to different users solving the prior art, to cannot provide for different types of crowd
The technical issues of more targetedly servicing.
To achieve the above object, it the present invention provides a kind of user group's classification method, the described method comprises the following steps:
The user data and device data of acquisition user to be sorted is buried by the way that client is preset；
Extract equipment identifies from the device data, searches whether exist and the device identification in user tag library
Associated historical user's label；
If it does not exist, then the corresponding target user's label of the user to be sorted is obtained according to the user data；
Target user's label is input to default crowd's Definition Model, and according to model export result determine it is described to
Target group's type belonging to sorted users.
Preferably, the step that the corresponding target user's label of the user to be sorted is obtained according to the user data
Suddenly, comprising:
Read the user attribute data for including in the user data and user behavior data；
Tag definition is carried out to the user attribute data based on the attribute tags database constructed in advance, user is obtained and belongs to
Property label；
Tag definition is carried out to the user behavior data based on the behavior tag database constructed in advance, obtains user's row
For label；
Using the user property label and the user behavior label as the corresponding target user of the user to be sorted
Label.
Preferably, described fixed to user attribute data progress label based on the attribute tags database constructed in advance
The step of justice, acquisition user property label, comprising:
The user attribute data is divided into build-in attribute, Asset Attributes and credit attribute by preset attribute dimension；
Attribution rule matching is carried out to the build-in attribute, the Asset Attributes and the credit attribute respectively, to obtain
What the assets rule of the intrinsic rule, Asset Attributes satisfaction that take the build-in attribute to meet and the credit attribute met
Credit rule；
It is regular and described to search the intrinsic rule, the assets respectively in the attribute tags database constructed in advance
The corresponding label of credit rule, and using the label found as the user property label of the user to be sorted.
Preferably, described fixed to user behavior data progress label based on the behavior tag database constructed in advance
The step of justice, acquisition user behavior label, comprising:
Behavior keyword and behavior frequency data are read from the user behavior data；
Rule of conduct matching is carried out to the behavior keyword and the behavior frequency data respectively, to obtain the row
For the corresponding goal behavior rule of keyword and the corresponding target frequency rule of the frequency data；
Search the goal behavior rule and target frequency rule respectively in the behavior tag database constructed in advance
Then corresponding label, and using the label found as the user behavior label of the user to be sorted.
Preferably, described that target user's label is input to default crowd's Definition Model, and exported and tied according to model
Fruit determines the step of target group's type belonging to the user to be sorted, comprising:
Target user's label is input to default crowd's Definition Model so that default crowd's Definition Model according to
It includes any one of described target user's label or multinomial that target user's label is searched in crowd's tag database
The crowd of label identifies；
It is corresponding that crowd's mark is searched in mapping relations between the crowd's mark constructed in advance and crowd's type
Target group's type.
Preferably, the extract equipment from device data mark, search whether to exist in user tag library with
After the step of device identification associated historical user's label, the method also includes:
If it exists, then historical user's label is obtained, and the user couple to be sorted is obtained according to the user data
The target user's label answered；
Detect whether historical user's label is contained in target user's label, if it is not, then to the historical user
Label and target user's label are combined, to obtain combination tag collection；
The combination tag collection is input to default crowd's Definition Model, and determining described wait divide according to model output result
Target group's type belonging to class user.
Preferably, described that the step of result determines target group's type belonging to the user to be sorted is exported according to model
Later, the method also includes:
The corresponding user preferences of target group's type are searched in user preference data library based on fuzzy matching algorithm；
The corresponding information to be pushed of the user preferences is obtained, the information to be pushed is carried out according to the user data
Priority ranking, and first information to be pushed of sorting pushes to the user to be sorted.
In addition, to achieve the above object, the present invention also proposes a kind of user group's sorter, described device includes:
Data acquisition module, for burying the user data and number of devices of acquisition user to be sorted by the way that client is preset
According to；
Label lookup module searches whether in user tag library for the extract equipment mark from the device data
In the presence of historical user's label associated with the device identification；
Label acquisition module, for there is no historical users associated with the device identification to mark in user tag library
When label, the corresponding target user's label of the user to be sorted is obtained according to the user data；
Crowd's determining module, for target user's label to be input to default crowd's Definition Model, and according to model
Output result determines target group's type belonging to the user to be sorted.
In addition, to achieve the above object, the present invention also proposes a kind of user group's sorting device, the equipment includes: to deposit
Reservoir, processor and it is stored in the user group's sort program that can be run on the memory and on the processor, it is described
User group's sort program is arranged for carrying out the step of user group's classification method as described above.
In addition, to achieve the above object, the present invention also proposes a kind of storage medium, user is stored on the storage medium
Group classification program, user group's sort program realize user group classification side as described above when being executed by processor
The step of method.
The present invention buries the user data and device data of acquisition user to be sorted by the way that client is preset；From number of devices
It is identified according to middle extract equipment, searches whether there is historical user's label associated with device identification in user tag library；If
There is no then obtain the corresponding target user's label of user to be sorted according to user data；Target user's label is input to default
Crowd's Definition Model, and result is exported according to model and determines target group's type belonging to user to be sorted, since the present invention is
The corresponding historical user's label of user is first searched according to collected device data, is continued if not finding according to user data
Tag definition is carried out to user and obtains user tag, the label that then will acquire again inputs crowd's Definition Model and according to model
It exports result and finally determines crowd's type belonging to user, be subsequent to realize the accurate positionin to user crowd's type
Pertinent service is carried out to user and provides effective foundation.
Detailed description of the invention
Fig. 1 is the structural representation of the user group's sorting device for the hardware running environment that the embodiment of the present invention is related to
Figure；
Fig. 2 is the flow diagram of user group's classification method first embodiment of the present invention；
Fig. 3 is the flow diagram of user group's classification method second embodiment of the present invention；
Fig. 4 is the flow diagram of user group's classification method 3rd embodiment of the present invention；
Fig. 5 is the structural block diagram of user group's sorter first embodiment of the present invention.
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific embodiment
It should be appreciated that described herein, specific examples are only used to explain the present invention, is not intended to limit the present invention.
Referring to Fig.1, Fig. 1 is the user group's sorting device structure for the hardware running environment that the embodiment of the present invention is related to
Schematic diagram.
As shown in Figure 1, user group's sorting device may include: processor 1001, such as central processing unit
(Central Processing Unit, CPU), communication bus 1002, user interface 1003, network interface 1004, memory
1005.Wherein, communication bus 1002 is for realizing the connection communication between these components.User interface 1003 may include display
Shield (Display), input unit such as keyboard (Keyboard), optional user interface 1003 can also include that the wired of standard connects
Mouth, wireless interface.Network interface 1004 optionally may include standard wireline interface and wireless interface (such as Wireless Fidelity
(WIreless-FIdelity, WI-FI) interface).Memory 1005 can be the random access memory (Random of high speed
Access Memory, RAM) memory, be also possible to stable nonvolatile memory (Non-Volatile Memory,
), such as magnetic disk storage NVM.Memory 1005 optionally can also be the storage device independently of aforementioned processor 1001.
It will be understood by those skilled in the art that structure shown in Fig. 1 does not constitute the limit to user group's sorting device
It is fixed, it may include perhaps combining certain components or different component layouts than illustrating more or fewer components.
As shown in Figure 1, as may include operating system, data storage mould in a kind of memory 1005 of storage medium
Block, network communication module, Subscriber Interface Module SIM and user group's sort program.
In user group's sorting device shown in Fig. 1, network interface 1004 is mainly used for being counted with network server
According to communication；User interface 1003 is mainly used for carrying out data interaction with user；Processing in user group's sorting device of the present invention
Device 1001, memory 1005 can be set in user group's sorting device, and user group's sorting device passes through processor
The user group's sort program stored in 1001 calling memories 1005, and execute user group provided in an embodiment of the present invention point
Class method.
The embodiment of the invention provides a kind of user group's classification methods, are user group of the present invention point referring to Fig. 2, Fig. 2
The flow diagram of class method first embodiment.
In the present embodiment, user group's classification method the following steps are included:
Step S10: the user data and device data of acquisition user to be sorted is buried by the way that client is preset；
It should be noted that the executing subject of the method for the present invention can be data management platform (Data Management
Platform, DMP platform), i.e., a kind of system that the multiparty data of dispersion can be put under the uniform management is acquired data, is whole
Splitting or integrating class, analysis, distribution platform.
It should be understood that the user data may include all operation datas that user inputs in the client, with
And the account data that user inputs in login or registered client account.It is corresponding hard that the device data can be client
Part facility information.It is so-called to bury a little, i.e., by way of code is set in the corresponding carrier of target data source and carries out data acquisition.
In the concrete realization, DMP platform can be by acquiring use to be sorted in burying a little for client-side setting in advance in real time
The user data and device data at family.
Step S20: from the device data extract equipment identify, searched whether in user tag library exist with it is described
The associated historical user's label of device identification；
It should be noted that the device identification can be similar to device hardware identification code, equipment factory code, international shifting
Dynamic device identification (International Mobile Equipment Identity, IMEI) etc. can distinguish distinct device or
Characterize the mark character of equipment uniqueness.
It should be understood that user tag can be used to characterize the personal characteristics of different consumers or user, consumption propensity or
The personalization attributes such as hobby.In practical applications, the user that may currently log in is new registration user；It is also possible to currently log in
User be not new registration user, only replaced the terminal device of Account Logon；It is also possible that chartered old use
The user is marked according to user data when logging in family and DMP platform is in user's first time logon account
Signing justice, and user tag data are incorporated after the label after definition is associated with the terminal device of account and Account Logon
Library.
In order to be accurately distinguished to above-mentioned three kinds of situations, in the present embodiment DMP platform by bury a little get to minute
After the device data of class user, will also the device identification for including in device data be read, then in the user tag constructed in advance
Historical user's label associated with the device identification is searched in database, to determine whether the user is user not to be undergone to mark
Sign the new user of justice.
Certainly, in order to more accurately judge whether a certain user is the new user, DMP platform will also be stepped on according to user
The account of record is further detected, that is, whether checks the corresponding account identification of the user account (Identification, ID)
It is associated with corresponding historical user's label；Determine that it is not belonging to new user if having, if without determining that it belongs to new user.
Step S30: if it does not exist, then the corresponding target user of the user to be sorted is obtained according to the user data and marked
Label；
It should be noted that the user data can be divided into user behavior data in the present embodiment and following each embodiments
And user attribute data.Wherein, the user behavior data includes but is not limited to behavior on line, marketing behavior and business conduct
Deng；By taking real estate industry as an example, behavior includes: search key, browsing source of houses etc. on the line, and business conduct includes purchase, report
Name sees room, participation activity, outgoing call behavior, intention feedback etc.；Marketing behavior includes feedback, the intention etc. to marketing activity.It is described
User attribute data can be divided into: build-in attribute, Asset Attributes and credit attribute；Wherein, the build-in attribute includes gender, year
Age, occupation, region, educational background, marital status etc., the Asset Attributes include income level, house situation, vehicle condition, loan
Situation etc., the credit attribute include credit card amount, credit card repayment situation, reference situation etc..
In the concrete realization, there is no associated with the device identification in determining user tag library for DMP platform
When historical user's label, the corresponding target user's label of user to be sorted can be obtained according to user data.
Certainly, before executing this step, research staff can construct one for carrying out tag definition to user in advance
Label system, the label system may include three classes level-one label, such as intrinsic label, frequency label and combination tag；Each level-one
Label may include one or more second level labels again, such as intrinsic label may include the second levels labels such as build-in attribute, intrinsic behavior；
Frequency label include the time range frequency and how many days in the second levels label such as the frequency, and the corresponding second level label of combination tag then may be used
It is made of the intersecting and merging of a variety of intrinsic labels or frequency label, benefit.In addition, the definition of all kinds of labels is with user data (such as property
Not, No. x room, the x month-y months y are seen in age, region, income level, the level of consumption, the browsing source of houses, search key, purchase, registration
Between number in the city z registration see adopt a young relative, interior had house-purchase intention etc. in the city m within nearest n days) based on.
Step S40: target user's label is input to default crowd's Definition Model, and true according to model output result
Target group's type belonging to the fixed user to be sorted.
It should be noted that default crowd's Definition Model can be a kind of Data searching model, it is also possible to one kind
Data Matching model, default crowd's Definition Model described in the present embodiment can be looked into corresponding database according to user tag
It looks for all comprising the label or crowd's type with the label characteristics.For example, DMP platform is fixed according to the user data of user A
Justice goes out the label of user A are as follows: attribute tags " 23 years old, undergraduate course ", behavior label " browse rent a house list page, number 50 in nearest one week
It is secondary ", after these labels are input to crowd's Definition Model, crowd's Definition Model can substantially determine user according to lookup result
Crowd's type of A is " graduating student, intention of renting a house user ", that is, has the graduating student for intention of renting a house.
In the concrete realization, target user's label can be input to default crowd's Definition Model by DMP platform, so that institute
Stating default crowd's Definition Model and being searched in crowd's tag database according to target user's label includes that the target is used
The crowd of any one of family label or multinomial label mark；Mapping between the crowd's mark constructed in advance and crowd's type
The crowd is searched in relationship identifies corresponding target group's type.Wherein, crowd's mark can be for characterizing difference
The identification information of crowd, such as " student=student ", " youth=young people ", " tenant=tenant " " high frequency
Rate=altofrequency " etc..
Further, it is the extension service for realizing fining, meets the different demands of different user.DMP in the present embodiment
Platform, can also be based on fuzzy matching algorithm in user preference data after determining target group's type belonging to user to be sorted
The corresponding user preferences of target group's type are searched in library；The corresponding information to be pushed of the user preferences is obtained, according to
The user data carries out priority ranking to the information to be pushed, and by sort first information to be pushed push to it is described
User to be sorted.Wherein, the fuzzy matching algorithm, also referred to as fuzzy search algorithm allow searched information and search to put question to
Between the search pattern that has a certain difference.The user preference data library can be according to all kinds of same people type of user
Shared hobby building database.
For example, DMP platform is determining that target group's type belonging to user A is the " graduate of the current year with intention of renting a house
It is raw " after, the user preferences that this kind of crowd is searched in user preference data library are " low rent, source of houses position and environmental requirement phase
/ whole group is shared to lower, traffic convenience, the mode of renting a house ", it can search for according to above-mentioned user preferences DMP platform corresponding
Information of real estate is then based on the information of real estate searched according to the occupational information (such as CompanyAddress) in user data to the source of houses
Information carries out the priority ranking of trip distance distance, then source of houses place and CompanyAddress is preferential apart from shortest information of real estate
Recommend user A.
The present embodiment buries the user data and device data of acquisition user to be sorted by the way that client is preset；From equipment
Extract equipment identifies in data, searches whether there is historical user's label associated with device identification in user tag library；
The corresponding target user's label of user to be sorted is then obtained according to user data if it does not exist；Target user's label is input to pre-
If crowd's Definition Model, and result is exported according to model and determines target group's type belonging to user to be sorted, due to this implementation
Example is first to search the corresponding historical user's label of user according to collected device data, is continued if not finding according to user
Data carry out tag definition to user and obtain user tag, the label input crowd's Definition Model and basis that then will acquire again
Model output result finally determines that crowd's type belonging to user is to realize the accurate positionin to user crowd's type
It is subsequent that effective foundation is provided to user's progress pertinent service.
With reference to Fig. 3, Fig. 3 is the flow diagram of user group's classification method second embodiment of the present invention.
Based on above-mentioned first embodiment, in the present embodiment, the step S30 includes:
Step S301: the user attribute data for including in the user data and user behavior data are read；
It should be understood that the user data can be divided into user behavior data and user attribute data in the present embodiment.
Wherein, the user behavior data includes but is not limited to behavior on line, marketing behavior and business conduct etc.；It is with real estate industry
Example, behavior includes: search key, browsing source of houses etc. on the line, business conduct include purchase, registration see room, participation activity,
Outgoing call behavior, intention feedback etc.；Marketing behavior includes feedback, the intention etc. to marketing activity.The user attribute data can divide
Are as follows: build-in attribute, Asset Attributes and credit attribute；Wherein, the build-in attribute include gender, the age, occupation, region, educational background,
Marital status etc., the Asset Attributes include income level, house situation, vehicle condition, loan profile etc., the credit attribute
Including credit card amount, credit card repayment situation, reference situation etc..
In the concrete realization, DMP platform can be read from collected user data to be sorted according to above-mentioned data dimension
The corresponding user attribute data of user and user behavior data.
Step S302: carrying out tag definition to the user attribute data based on the attribute tags database constructed in advance,
Obtain user property label；
It should be noted that research staff can targetedly establish an attribute mark based on the label system constructed in advance
Database is signed, stores the corresponding relationship having between " attribute-rule-label ", such as the category of user A in the attribute tags database
Property is that " 24 years old, university student, undergraduate course " then its corresponding rule is " student-university student-(student at school, the preceding-year-pupil, graduate of the current year
It is raw) ", the user property label after being matched according to rule is " graduating student ".
Specifically, the user attribute data can be divided into build-in attribute, assets category by preset attribute dimension by DMP platform
Property and credit attribute；Then attribute rule are carried out to the build-in attribute, the Asset Attributes and the credit attribute respectively
It then matches, to obtain intrinsic rule, the assets rule of Asset Attributes satisfaction and the letter that the build-in attribute meets
The credit rule met with attribute；Search the intrinsic rule, the money respectively in the attribute tags database constructed in advance
Rule and the corresponding label of credit rule are produced, and using the label found as the user property of the user to be sorted
Label.
Step S303: carrying out tag definition to the user behavior data based on the behavior tag database constructed in advance,
Obtain user behavior label；
It should be noted that research staff can targetedly establish a behavior mark based on the label system constructed in advance
Database is signed, stores the corresponding relationship having between " behavior-rule-label ", such as the row of user A in behavior tag database
To be that " nearest one week browsing rent a house list page, number 50 times " then its corresponding rule is " frequency rule-how many days in the frequency ",
User property label after being matched according to rule is " intention of renting a house user ".
Specifically, DMP platform can be read from the user behavior data behavior keyword (as browsing new house, second-hand house,
List page of renting a house or down payment, band has seen room etc.) and behavior frequency data (between such as 5.1-5.7 Shanghai registration seen
Room, it is 5 days nearest in Beijing had house-purchase intention etc.)；Respectively to the behavior keyword and the behavior frequency data into
Every trade is rule match, to obtain the corresponding goal behavior rule of the behavior keyword and the corresponding target of the frequency data
Frequency rule；Search the goal behavior rule and target frequency rule respectively in the behavior tag database constructed in advance
Then corresponding label, and using the label found as the user behavior label of the user to be sorted.
Step S304: the user property label and the user behavior label is corresponding as the user to be sorted
Target user's label.
In the concrete realization, DMP platform is in the user property label and use for defining user to be sorted according to user data
After family behavior label, these labels that can be will acquire are collectively as the corresponding target user's label of user to be sorted.
In addition, it should be noted that, the label of user be not it is unalterable, can do according to the actual situation dynamic adjust
It is whole, such as after user A graduates work 1 year, DMP platform can be according to its job information to its " graduate of the current year predetermined
Life " label is adjusted.
Further, in this embodiment the user tag that after getting user tag, can also will acquire of DMP platform with
The account that user to be sorted currently logs in is bound, and specifically be can be and is bound account ID with user tag, can also be with
The device identification of all registration terminals is bound with user identifier when being by Account Logon, and the present embodiment does not limit this
System.
The present embodiment is by reading the user attribute data for including in user data and user behavior data；Based on preparatory
The attribute tags database of building carries out tag definition to user attribute data, obtains user property label；Based on preparatory building
Behavior tag database to user behavior data carry out tag definition, obtain user behavior label；By user property label and
User behavior label is as the corresponding target user's label of user to be sorted, due to being two dimensions of behavior and attribute from user
To carry out tag definition to user, to ensure that the accuracy that user tag defines.
With reference to Fig. 4, Fig. 4 is the flow diagram of user group's classification method 3rd embodiment of the present invention.
Based on the various embodiments described above, in the present embodiment, after the step S40, the method also includes:
Step S50: if it exists, then historical user's label is obtained, and described wait divide according to user data acquisition
The corresponding target user's label of class user；
It will be appreciated that if there is go through associated with the device identification in DMP platform in detecting user tag library
When history user tag, then determining that the user to be sorted is not is new registration user, can obtain the historical user of the user at this time
Then label executes step in above-mentioned second embodiment further according to burying a little collected user data, to obtain use to be sorted
The corresponding target user's label in family.
It should be understood that may be deposited in target user's label that DMP platform is obtained according to current collected user data
In the shared label intersected with historical user's label, it is also possible to which there is no shared labels, such as user A is after renting a house 1 year
Certain time abruptly starts to the building or information of real estate of frequently paying close attention to new house, second-hand house, then DMP platform is in advance to user A
Historical user's label of definition may then need further progress to adjust, and such as be adjusted to from " tenant " " Intention of purchase housing person ".
Step S60: detecting whether historical user's label is contained in target user's label, if it is not, then to described
Historical user's label and target user's label are combined, to obtain combination tag collection；
In the concrete realization, DMP platform also will test historical user's mark predetermined after obtaining target user's label
Whether label are included in target user's label that this is obtained, if then according to target user's label to the historical user
Label is replaced；If it does not exist, then it whether there is same nature in first detection history user tag and target user's label
Label (such as tenant and house purchaser, students and teacher, lawyer, office worker, civil servant etc.), according to target user's label if having
In label the user tag of same nature in historical user's label is replaced；To historical user's label if not
It is combined with target user's label, to obtain combination tag collection.
Step S70: the combination tag collection is input to default crowd's Definition Model, and result is exported according to model and is determined
Target group's type belonging to the user to be sorted.
In the concrete realization, DMP platform is after getting combination tag collection, and all labels that can concentrate combination tag are all
It is input to default crowd's Definition Model, result is then exported according to model and determines target of the user to be sorted belonging to present period
Crowd's type.
There is historical user's label associated with device identification in detecting user tag library in the present embodiment DMP platform
When, then historical user's label is obtained, and the corresponding target user's label of user to be sorted is obtained according to user data；Detection history
Whether user tag is contained in target user's label, if being otherwise combined to historical user's label and target user's label, with
Obtain combination tag collection；Combination tag collection is input to default crowd's Definition Model, and result is exported according to model and is determined wait divide
Target group's type belonging to class user, due to being that the target user's label by historical user's label and currently obtained is combined
Target group's type belonging to user is determined afterwards, so that the crowd's type finally determined is more accurate.
In addition, the embodiment of the present invention also proposes a kind of storage medium, user group's classification is stored on the storage medium
Program, user group's sort program realize the step of user group's classification method as described above when being executed by processor
Suddenly.
It is the structural block diagram of user group's sorter first embodiment of the present invention referring to Fig. 5, Fig. 5.
As shown in figure 5, user group's sorter that the embodiment of the present invention proposes includes:
Data acquisition module 501, for by the preset user data for burying acquisition user to be sorted of client and setting
Standby data；
Label lookup module 502, for the extract equipment mark from the device data, searching in user tag library is
It is no to there is historical user's label associated with the device identification；
Label acquisition module 503, for there is no history associated with the device identification to use in user tag library
When the label of family, the corresponding target user's label of the user to be sorted is obtained according to the user data；
Crowd's determining module 504, for target user's label to be input to default crowd's Definition Model, and according to mould
Type output result determines target group's type belonging to the user to be sorted.
The present embodiment buries the user data and device data of acquisition user to be sorted by the way that client is preset；From equipment
Extract equipment identifies in data, searches whether there is historical user's label associated with device identification in user tag library；
The corresponding target user's label of user to be sorted is then obtained according to user data if it does not exist；Target user's label is input to pre-
If crowd's Definition Model, and result is exported according to model and determines target group's type belonging to user to be sorted, due to this implementation
Example is first to search the corresponding historical user's label of user according to collected device data, is continued if not finding according to user
Data carry out tag definition to user and obtain user tag, the label input crowd's Definition Model and basis that then will acquire again
Model output result finally determines that crowd's type belonging to user is to realize the accurate positionin to user crowd's type
It is subsequent that effective foundation is provided to user's progress pertinent service.
Based on the above-mentioned user group's sorter first embodiment of the present invention, user group's sorter of the present invention is proposed
Second embodiment.
In the present embodiment, the label acquisition module 503 is also used to read the user in the user data included and belongs to
Property data and user behavior data；Label is carried out to the user attribute data based on the attribute tags database constructed in advance
Definition obtains user property label；Label is carried out to the user behavior data based on the behavior tag database constructed in advance
Definition obtains user behavior label；Using the user property label and the user behavior label as the user to be sorted
Corresponding target user's label.
Further, the label acquisition module 503 is also used to draw the user attribute data by preset attribute dimension
It is divided into build-in attribute, Asset Attributes and credit attribute；Respectively to the build-in attribute, the Asset Attributes and the credit
Attribute carries out attribution rule matching, to obtain intrinsic rule, the assets of Asset Attributes satisfaction that the build-in attribute meets
The credit rule that the regular and described credit attribute meets；It is searched respectively in the attribute tags database constructed in advance described solid
Regular, the described assets rule and the corresponding label of credit rule, and using the label found as described to be sorted
The user property label of user.
Further, the label acquisition module 503 is also used to read behavior keyword from the user behavior data
And behavior frequency data；Rule of conduct matching is carried out to the behavior keyword and the behavior frequency data respectively, with
Obtain the corresponding goal behavior rule of the behavior keyword and the corresponding target frequency rule of the frequency data；In preparatory structure
The goal behavior rule and the corresponding label of target frequency rule are searched in the behavior tag database built respectively, and will
User behavior label of the label found as the user to be sorted.
Further, crowd's determining module 504 is also used to target user's label being input to default crowd and determine
Adopted model, so that default crowd's Definition Model is searched in crowd's tag database according to target user's label and includes
There is the crowd of any one of described target user's label or multinomial label mark；In the crowd's mark and crowd's class constructed in advance
The crowd is searched in mapping relations between type identifies corresponding target group's type.
Further, user group's sorter provided in this embodiment further includes tag combination module, the set of tags
Block being molded, when for there is historical user's label associated with the device identification in user tag library, being gone through described in acquisition
History user tag, and the corresponding target user's label of the user to be sorted is obtained according to the user data；It is gone through described in detection
Whether history user tag is contained in target user's label, if it is not, then to historical user's label and the target user
Label is combined, to obtain combination tag collection；The combination tag collection is input to default crowd's Definition Model, and according to mould
Type output result determines target group's type belonging to the user to be sorted.
Further, user group's sorter provided in this embodiment further includes info push module, and the information pushes away
Module is sent, for searching the corresponding user's happiness of target group's type in user preference data library based on fuzzy matching algorithm
It is good；The corresponding information to be pushed of the user preferences is obtained, the information to be pushed is carried out according to the user data preferential
Grade sequence, and first information to be pushed of sorting pushes to the user to be sorted.
The other embodiments or specific implementation of user group's sorter of the present invention can refer to above-mentioned each method and implement
Example, details are not described herein again.
It should be noted that, in this document, the terms "include", "comprise" or its any other variant are intended to non-row
His property includes, so that the process, method, article or the system that include a series of elements not only include those elements, and
And further include other elements that are not explicitly listed, or further include for this process, method, article or system institute it is intrinsic
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including being somebody's turn to do
There is also other identical elements in the process, method of element, article or system.
The serial number of the above embodiments of the invention is only for description, does not represent the advantages or disadvantages of the embodiments.
Through the above description of the embodiments, those skilled in the art can be understood that above-described embodiment side
Method can be realized by means of software and necessary general hardware platform, naturally it is also possible to by hardware, but in many cases
The former is more preferably embodiment.Based on this understanding, technical solution of the present invention substantially in other words does the prior art
The part contributed out can be embodied in the form of software products, which is stored in a storage medium
In (such as read-only memory/random access memory, magnetic disk, CD), including some instructions are used so that a terminal device (can
To be mobile phone, computer, server, air conditioner or the network equipment etc.) execute method described in each embodiment of the present invention.
The above is only a preferred embodiment of the present invention, is not intended to limit the scope of the invention, all to utilize this hair
Equivalent structure or equivalent flow shift made by bright specification and accompanying drawing content is applied directly or indirectly in other relevant skills
Art field, is included within the scope of the present invention.
Claims (10)
1. a kind of user group's classification method, which is characterized in that the described method includes:
The user data and device data of acquisition user to be sorted is buried by the way that client is preset；
Extract equipment identifies from the device data, searches whether to exist in user tag library related to the device identification
Historical user's label of connection；
If it does not exist, then the corresponding target user's label of the user to be sorted is obtained according to the user data；
Target user's label is input to default crowd's Definition Model, and according to model export result determine it is described to be sorted
Target group's type belonging to user.
2. the method as described in claim 1, which is characterized in that described to obtain the user to be sorted according to the user data
The step of corresponding target user's label, comprising:
Read the user attribute data for including in the user data and user behavior data；
Tag definition is carried out to the user attribute data based on the attribute tags database constructed in advance, obtains user property mark
Label；
Tag definition is carried out to the user behavior data based on the behavior tag database constructed in advance, obtains user behavior mark
Label；
Using the user property label and the user behavior label as the corresponding target user's label of the user to be sorted.
3. method according to claim 2, which is characterized in that it is described based on the attribute tags database constructed in advance to described
The step of user attribute data carries out tag definition, obtains user property label, comprising:
The user attribute data is divided into build-in attribute, Asset Attributes and credit attribute by preset attribute dimension；
Attribution rule matching is carried out to the build-in attribute, the Asset Attributes and the credit attribute respectively, to obtain
State the intrinsic rule of build-in attribute satisfaction, the assets rule that the Asset Attributes meet and the credit that the credit attribute meets
Rule；
Search the intrinsic rule, assets rule and the credit respectively in the attribute tags database constructed in advance
The corresponding label of rule, and using the label found as the user property label of the user to be sorted.
4. method as claimed in claim 3, which is characterized in that it is described based on the behavior tag database constructed in advance to described
The step of user behavior data carries out tag definition, obtains user behavior label, comprising:
Behavior keyword and behavior frequency data are read from the user behavior data；
Rule of conduct matching is carried out to the behavior keyword and the behavior frequency data respectively, is closed with obtaining the behavior
The corresponding goal behavior rule of keyword and the corresponding target frequency rule of the frequency data；
It searches the goal behavior rule respectively in the behavior tag database constructed in advance and the target frequency rule is right
The label answered, and using the label found as the user behavior label of the user to be sorted.
5. the method as described in claim 1, which is characterized in that described target user's label is input to default crowd to determine
Adopted model, and the step of result determines target group's type belonging to the user to be sorted is exported according to model, comprising:
Target user's label is input to default crowd's Definition Model, so that default crowd's Definition Model is according to
It includes any one of described target user's label or multinomial label that target user's label is searched in crowd's tag database
Crowd mark；
The crowd, which is searched, in mapping relations between the crowd's mark constructed in advance and crowd's type identifies corresponding target
Crowd's type.
6. such as method described in any one of claim 1 to 5, which is characterized in that the extract equipment from the device data
It identifies, is searched whether in user tag library after there is the step of historical user's label associated with the device identification,
The method also includes:
If it exists, then historical user's label is obtained, and corresponding according to the user data acquisition user to be sorted
Target user's label；
Detect whether historical user's label is contained in target user's label, if it is not, then to historical user's label
It is combined with target user's label, to obtain combination tag collection；
The combination tag collection is input to default crowd's Definition Model, and result is exported according to model and determines the use to be sorted
Target group's type belonging to family.
7. method as claimed in claim 6, which is characterized in that described to determine the user to be sorted according to model output result
After the step of affiliated target group's type, the method also includes:
The corresponding user preferences of target group's type are searched in user preference data library based on fuzzy matching algorithm；
The corresponding information to be pushed of the user preferences is obtained, the information to be pushed is carried out according to the user data preferential
Grade sequence, and first information to be pushed of sorting pushes to the user to be sorted.
8. a kind of user group's sorter, which is characterized in that described device includes:
Data acquisition module, for burying the user data and device data of acquisition user to be sorted by the way that client is preset；
Label lookup module searches whether exist in user tag library for the extract equipment mark from the device data
Historical user's label associated with the device identification；
Label acquisition module, for historical user's label associated with the device identification to be not present in user tag library
When, the corresponding target user's label of the user to be sorted is obtained according to the user data；
Crowd's determining module for target user's label to be input to default crowd's Definition Model, and is exported according to model
As a result target group's type belonging to the user to be sorted is determined.
9. a kind of user group's sorting device, which is characterized in that the equipment includes: memory, processor and is stored in described
On memory and the user group's sort program that can run on the processor, user group's sort program are configured to reality
Now the step of user group's classification method as described in any one of claims 1 to 7.
10. a kind of storage medium, which is characterized in that be stored with user group's sort program, the user on the storage medium
Group classification program realizes the step of user group's classification method as described in any one of claim 1 to 7 when being executed by processor
Suddenly.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910046571.1A CN109919652A (en) | 2019-01-17 | 2019-01-17 | User group's classification method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910046571.1A CN109919652A (en) | 2019-01-17 | 2019-01-17 | User group's classification method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109919652A true CN109919652A (en) | 2019-06-21 |
Family
ID=66960326
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910046571.1A Pending CN109919652A (en) | 2019-01-17 | 2019-01-17 | User group's classification method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109919652A (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110335078A (en) * | 2019-07-03 | 2019-10-15 | 中国银行股份有限公司 | Advertisement sending method, device and equipment |
| CN110503453A (en) * | 2019-07-05 | 2019-11-26 | 平安银行股份有限公司 | Customer-action analysis method, apparatus, computer equipment and storage medium |
| CN110598769A (en) * | 2019-08-30 | 2019-12-20 | 京东数字科技控股有限公司 | User group discovery method, device, equipment and computer readable storage medium |
| CN110765101A (en) * | 2019-09-09 | 2020-02-07 | 湖南天云软件技术有限公司 | Label generation method and device, computer readable storage medium and server |
| CN110955690A (en) * | 2019-08-21 | 2020-04-03 | 广州云徙科技有限公司 | Self-service data labeling platform and self-service data labeling method based on big data technology |
| CN111178949A (en) * | 2019-12-18 | 2020-05-19 | 北京文思海辉金信软件有限公司 | Service resource matching reference data determination method, device, equipment and storage medium |
| CN111339402A (en) * | 2020-02-10 | 2020-06-26 | 口碑(上海)信息技术有限公司 | Service processing method and device |
| CN111400289A (en) * | 2020-02-23 | 2020-07-10 | 中国平安财产保险股份有限公司 | Intelligent user classification method, server and storage medium |
| CN111400599A (en) * | 2020-03-17 | 2020-07-10 | 苏宁金融科技(南京)有限公司 | User group portrait generation method, device and system |
| CN112215655A (en) * | 2020-10-13 | 2021-01-12 | 广东电网有限责任公司 | Client portrait label management method and system |
| CN112506981A (en) * | 2021-02-05 | 2021-03-16 | 深圳市阿卡索资讯股份有限公司 | Online training service pushing method and device |
| CN112506975A (en) * | 2020-11-30 | 2021-03-16 | 橙脑教育科技(上海)有限公司 | Information recommendation method, device and system |
| CN113626708A (en) * | 2021-08-11 | 2021-11-09 | 武汉卓尔数字传媒科技有限公司 | Service recommendation method and device, electronic equipment and storage medium |
| CN113762994A (en) * | 2020-06-08 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Method and device for user operation management |
| CN114912946A (en) * | 2022-04-24 | 2022-08-16 | 零犀(北京)科技有限公司 | Method, device, storage medium and electronic equipment for determining user hierarchy |
| CN115357767A (en) * | 2022-08-23 | 2022-11-18 | 晋商消费金融股份有限公司 | User label portrait method and system |
| CN115600600A (en) * | 2022-10-26 | 2023-01-13 | 中电金信软件有限公司（Cn） | Label naming method and device of multi-object label system, electronic equipment and medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106503015A (en) * | 2015-09-07 | 2017-03-15 | 国家计算机网络与信息安全管理中心 | A kind of method for building user's portrait |
| CN106776716A (en) * | 2016-11-21 | 2017-05-31 | 北京齐尔布莱特科技有限公司 | A kind of intelligent Matching marketing consultant and the method and apparatus of user |
| CN107086922A (en) * | 2016-02-15 | 2017-08-22 | 中国移动通信集团福建有限公司 | A kind of user behavior recognition method and apparatus |
| CN108230113A (en) * | 2018-01-16 | 2018-06-29 | 平安好房（上海）电子商务有限公司 | User's portrait generation method, device, equipment and readable storage medium storing program for executing |
| CN108846687A (en) * | 2018-04-02 | 2018-11-20 | 平安科技（深圳）有限公司 | Client segmentation method, apparatus and storage medium |
-
2019
- 2019-01-17 CN CN201910046571.1A patent/CN109919652A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106503015A (en) * | 2015-09-07 | 2017-03-15 | 国家计算机网络与信息安全管理中心 | A kind of method for building user's portrait |
| CN107086922A (en) * | 2016-02-15 | 2017-08-22 | 中国移动通信集团福建有限公司 | A kind of user behavior recognition method and apparatus |
| CN106776716A (en) * | 2016-11-21 | 2017-05-31 | 北京齐尔布莱特科技有限公司 | A kind of intelligent Matching marketing consultant and the method and apparatus of user |
| CN108230113A (en) * | 2018-01-16 | 2018-06-29 | 平安好房（上海）电子商务有限公司 | User's portrait generation method, device, equipment and readable storage medium storing program for executing |
| CN108846687A (en) * | 2018-04-02 | 2018-11-20 | 平安科技（深圳）有限公司 | Client segmentation method, apparatus and storage medium |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110335078A (en) * | 2019-07-03 | 2019-10-15 | 中国银行股份有限公司 | Advertisement sending method, device and equipment |
| CN110503453A (en) * | 2019-07-05 | 2019-11-26 | 平安银行股份有限公司 | Customer-action analysis method, apparatus, computer equipment and storage medium |
| CN110955690A (en) * | 2019-08-21 | 2020-04-03 | 广州云徙科技有限公司 | Self-service data labeling platform and self-service data labeling method based on big data technology |
| CN110598769B (en) * | 2019-08-30 | 2022-06-07 | 京东科技控股股份有限公司 | User group discovery method, device, equipment and computer readable storage medium |
| CN110598769A (en) * | 2019-08-30 | 2019-12-20 | 京东数字科技控股有限公司 | User group discovery method, device, equipment and computer readable storage medium |
| CN110765101A (en) * | 2019-09-09 | 2020-02-07 | 湖南天云软件技术有限公司 | Label generation method and device, computer readable storage medium and server |
| CN110765101B (en) * | 2019-09-09 | 2022-08-02 | 天云软件技术有限公司 | Label generation method and device, computer readable storage medium and server |
| CN111178949A (en) * | 2019-12-18 | 2020-05-19 | 北京文思海辉金信软件有限公司 | Service resource matching reference data determination method, device, equipment and storage medium |
| CN111339402A (en) * | 2020-02-10 | 2020-06-26 | 口碑(上海)信息技术有限公司 | Service processing method and device |
| CN111400289A (en) * | 2020-02-23 | 2020-07-10 | 中国平安财产保险股份有限公司 | Intelligent user classification method, server and storage medium |
| CN111400289B (en) * | 2020-02-23 | 2023-09-29 | 中国平安财产保险股份有限公司 | Intelligent user classification method, server and storage medium |
| CN111400599A (en) * | 2020-03-17 | 2020-07-10 | 苏宁金融科技(南京)有限公司 | User group portrait generation method, device and system |
| CN113762994A (en) * | 2020-06-08 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Method and device for user operation management |
| CN112215655A (en) * | 2020-10-13 | 2021-01-12 | 广东电网有限责任公司 | Client portrait label management method and system |
| CN112506975A (en) * | 2020-11-30 | 2021-03-16 | 橙脑教育科技(上海)有限公司 | Information recommendation method, device and system |
| CN112506981A (en) * | 2021-02-05 | 2021-03-16 | 深圳市阿卡索资讯股份有限公司 | Online training service pushing method and device |
| CN113626708A (en) * | 2021-08-11 | 2021-11-09 | 武汉卓尔数字传媒科技有限公司 | Service recommendation method and device, electronic equipment and storage medium |
| CN114912946A (en) * | 2022-04-24 | 2022-08-16 | 零犀(北京)科技有限公司 | Method, device, storage medium and electronic equipment for determining user hierarchy |
| CN114912946B (en) * | 2022-04-24 | 2024-01-30 | 零犀(北京)科技有限公司 | Method and device for determining user layering, storage medium and electronic equipment |
| CN115357767A (en) * | 2022-08-23 | 2022-11-18 | 晋商消费金融股份有限公司 | User label portrait method and system |
| CN115357767B (en) * | 2022-08-23 | 2023-12-19 | 晋商消费金融股份有限公司 | User tag portrait method and system |
| CN115600600A (en) * | 2022-10-26 | 2023-01-13 | 中电金信软件有限公司（Cn） | Label naming method and device of multi-object label system, electronic equipment and medium |
| CN115600600B (en) * | 2022-10-26 | 2023-10-17 | 中电金信软件有限公司 | Label naming method, device, electronic equipment and medium of multi-object label system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109919652A (en) | User group's classification method, device, equipment and storage medium | |
| CN107515915B (en) | User identification association method based on user behavior data | |
| KR101797856B1 (en) | Method and system for artificial intelligence learning using messaging service and method and system for relaying answer using artificial intelligence | |
| US20020120516A1 (en) | Mobile marketing method, mobile marketing system, mobile marketing server, and associated user terminal, analysis terminal, and program | |
| CN102099803A (en) | Method and computer system for automatically answering natural language questions | |
| CN105744005A (en) | Client positioning and analyzing method and server | |
| CN107730310A (en) | Electronic installation, the method and storage medium for building Retail networks Rating Model | |
| CN111738785A (en) | Product selection method, system and storage medium | |
| CN111966866A (en) | Data asset management method and device | |
| CN112241489B (en) | Information pushing method, device, readable storage medium and computer equipment | |
| CN103189885A (en) | Server, information-management method, information-management program, and computer-readable recording medium with said program recorded thereon | |
| CN103250151A (en) | Server, information-anagement method, information-management program, and computer-readable recording medium with said program recorded thereon | |
| US11886556B2 (en) | Systems and methods for providing user validation | |
| CN108140055A (en) | Trigger application message | |
| CN108932646A (en) | User tag verification method, device and electronic equipment based on operator | |
| CN101425981A (en) | Information publishing system and method for publishing information according to mutual exclusive indication | |
| CN110197426A (en) | A kind of method for building up of credit scoring model, device and readable storage medium storing program for executing | |
| CN112910953B (en) | Business data pushing method and device and server | |
| JPWO2019244849A1 (en) | Post information extraction control device, post information extraction control program | |
| JP2022178650A (en) | Information processing device, information processing method, and information processing program | |
| CN105163270B (en) | The recognition methods of intelligent terminal location status and device | |
| KR102515539B1 (en) | Linked discount system and method based on travel destination sharing | |
| JP7390772B2 (en) | Visit information provision system, visit information provision method, and its program | |
| Tantothai et al. | mipMAP: A mobile application for proximate social network communication | |
| CN115208831B (en) | Request processing method, device, equipment 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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190621 |
|
| WD01 | Invention patent application deemed withdrawn after publication |