CN109783731A - A kind of customized information pushing method and system - Google Patents
A kind of customized information pushing method and system Download PDFInfo
- Publication number
- CN109783731A CN109783731A CN201910017478.8A CN201910017478A CN109783731A CN 109783731 A CN109783731 A CN 109783731A CN 201910017478 A CN201910017478 A CN 201910017478A CN 109783731 A CN109783731 A CN 109783731A
- Authority
- CN
- China
- Prior art keywords
- user
- data
- record data
- application record
- unfinished
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 71
- 238000012549 training Methods 0.000 claims description 37
- 230000003542 behavioural effect Effects 0.000 claims description 31
- 238000010801 machine learning Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 13
- 238000003860 storage Methods 0.000 claims description 13
- 238000007418 data mining Methods 0.000 abstract description 2
- 230000006399 behavior Effects 0.000 description 6
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000012795 verification Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 2
- 239000010931 gold Substances 0.000 description 2
- 229910052737 gold Inorganic materials 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 235000019640 taste Nutrition 0.000 description 2
- 206010052428 Wound Diseases 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention discloses a kind of customized information pushing method and system, which comprises chooses alternative unfinished application record data；The alternative unfinished application record data are matched with application record data are completed；When there is no matching result, the alternative unfinished application record data are determined as not completing application record data；The unfinished application record data are inputted into user's portrait label model and carry out operation, to export user's portrait label of the corresponding unfinished application record data；Customized information is pushed to user belonging to the unfinished application record data according to user portrait label, to significantly improve the value of data mining and promote efficiency, minimum is reduced to the puzzlement of user simultaneously, does not influence user experience, hence it is evident that increase the pushing efficiency of customized information.
Description
Technical field
This specification is related to Internet technical field more particularly to a kind of customized information pushing method and system.
Background technique
At this stage, as internet is universal, the push of customized information becomes a kind of general need.Customized information refers to specially
The information formulated for specific user or user group.In the application of realization, customized information can provide customization demand, improve
The user experience of product increases user's viscosity of product.For example, the customization marketing message based on user is exactly a kind of common
Customized information.
For example, generally requiring user by internet application credit financing product according to prompt and carrying out multiple steps
Input.For example, being that input handset, mailbox or social account are registered first；Then the finance for inputting personal information, being applied
Product etc.；Finally submit application.But the case where many users during which occur and just abandoned halfway, not completing application,
Such as without clicking " submitting application ".However the user of application is not completed for these, very likely to other credit financings
The customized information (such as preferential) of product etc. is interested, even more the key point of developing market, has to the strong of push customized information
Strong demand.
But in fact, on the one hand, since process steps process design defect, data transimission and storage may slip,
Cause some completed applications by error logging not complete, so that user is further mistakenly sent customized information,
The puzzlement for causing user reduces user experience, also has a greatly reduced quality to customized information pushing efficiency.
On the other hand, if only considering " completing application " and " not completing application " the two situations, and indistinguishably
Customized information is sent to the user of unfinished application, then some users can be made mistakenly to be sent and the basic nothing of its specific requirements
The customized information of pass still may cause the puzzlement of user, and customized information specific aim is not strong, and efficiency and conversion ratio are low.
Summary of the invention
In view of the above problems, this specification is proposed to overcome the above problem in order to provide one kind or at least be partially solved
The customized information pushing method and system of the above problem.
In a first aspect, this specification provides a kind of customized information pushing method characterized by comprising
Choose alternative unfinished application record data；By the alternative unfinished application record data and Shen is completed
Data please be record to be matched；When there is no matching result, the alternative unfinished application record data are determined as not complete
At application record data；The unfinished application record data input user portrait label model is subjected to operation, with output pair
Answer user's portrait label of the unfinished application record data；Remembered according to user portrait label to the unfinished application
It records user belonging to data and pushes customized information.
Second aspect, this specification provide a kind of customized information pushing system characterized by comprising
Alternative recording unit, for choosing alternative unfinished application record data；Data matching unit, being used for will be described
Alternative unfinished application record data are matched with application record data are completed；Determination unit, for ought not match
When as a result, the alternative unfinished application record data are determined as not completing application record data；Model arithmetic unit is used
In the unfinished application record data input user portrait label model is carried out operation, with the corresponding unfinished Shen of output
It please record user's portrait label of data；Push unit is used for the relevant user type of label of drawing a portrait according to the user, to institute
It states user belonging to unfinished application record data and pushes customized information.
The third aspect, this specification provide a kind of server, including processor and memory: the memory is for storing
The program of any of the above-described the method；The processor is configured to being realized for executing the program stored in the memory
The step of any of the above-described the method.
Fourth aspect, this specification embodiment provide a kind of computer readable storage medium, are stored thereon with computer journey
Sequence, when which is executed by processor the step of realization any of the above-described the method.
This specification said one or multiple technical solutions at least have following one or more technical effects:
In the technical solution for implementing this specification, alternative unfinished application record data are utilized and application is completed
Record data are matched, and determine unfinished application record data, then remember to unfinished application by user's portrait label model
Record data carry out sort operation and export corresponding user's portrait label, according to the relevant user type of user portrait label
The method for pushing customized information to user belonging to the unfinished application record data, can only be compared to traditional push mode
It just can be carried out popularization after obtaining user's registration, substantially increase the value of data mining and promote efficiency, while user is stranded
It disturbs and is reduced to minimum, do not influence user experience, significantly increase the pushing efficiency of customized information.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to this explanation
The limitation of book.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings:
Fig. 1 shows the flow chart of customized information pushing method according to an embodiment of the invention.
Fig. 2 shows the structural schematic diagrams of customized information pushing system according to an embodiment of the invention.
Fig. 3 shows the schematic diagram of the calculating equipment in an embodiment according to the present invention.
Specific embodiment
This specification technical solution is described in detail below by attached drawing and specific embodiment, it should be understood that this theory
Specific features in bright book embodiment and embodiment are the detailed description to this specification technical solution, rather than to this theory
The restriction of bright book technical solution, in the absence of conflict, technical characteristic in this specification embodiment and embodiment can be with
It is combined with each other.
The terms "and/or", only a kind of incidence relation for describing affiliated partner, indicates that there may be three kinds of passes
System, for example, A and/or B, can indicate: individualism A exists simultaneously A and B, these three situations of individualism B.In addition, herein
Middle character "/" typicallys represent the relationship that forward-backward correlation object is a kind of "or".
Embodiment
Referring to FIG. 1, the present invention passes through one embodiment of customized information pushing method, comprising:
Step S110 chooses alternative unfinished application record data.
Wherein, above-mentioned steps can choose alternative unfinished application record data from first database, and described first
The data of database purchase include the related data buried in each user's application process of acquisition by setting data.For example, can
To be buried a little by the way that data are arranged in client-side program, or data are set in the application page and bury mode a little, acquires each user
Behavioral data in application process, unfinished application record data alternately are stored into first database.
For data to be arranged in client-side program and bury a little, in advance in user's application process in client development process
It is related to business framework setting data on path and buries a little obtaining user and applying for that the client traffic that passes through of service traffics calls link
On the tracking data of total interface parameter and/or method parameter that are related to, acquire out each user Shen based on above-mentioned tracking data
Please during behavioral data.Here different types of feature can be arranged in behavioral data according to specific needs, such as in base
The residence time of this information page, project input sequence, check when clicking next step do not fill out item, identification card number format checking,
Cell-phone number format check, mailbox format check, Bank Account Number format check, individual privacy information page residence time, abandon application
The page, the verification of proof attachment and/or request times etc. where registration.These above-mentioned behavioral datas can be by client
Setting data are buried mode a little and are got in program, similarly, compared to client-side program, data only are arranged in the application page
The possible limitation of mode buried a little is higher, and the behavioral data type feature got is less, but it is of the invention to nor affect on realization
Method.
Further, in data structure, setting includes the second database that completed application record is collected in storage, with
And the first database of the unfinished user's application process record of storage alternatively.In general, the second database is when user Shen
By the database of record deposit when please complete, for example registration center is set or applies for the database at center, data content
It is worth larger also more important, therefore its second database and other databases are in the state being relatively isolated.And the first number
What it is according to library record is all users i.e. behavioral data of full dose user application process, i.e., does not complete the still complete of application either
At will collecting and recording for application.For example, unfinished application record can be stamped " unfinished " in first database
Label, but this label may be inaccuracy, therefore the completed application record recorded in the second database is by it
He carried out verification by approach, this process is different from the data of first database, and even the second database is by not in other words
Data with source are included.In this way, the data in first database need to be carried out according to the data in the second database
Matching verifying, and further screening is done,
Step S120 matches the alternative unfinished application record data with application record data are completed；
When there is no matching result, the alternative unfinished application record data are determined as not completing application record data.
Further, it is completed whether application record data include the alternative unfinished application record number described in judgement
User Identity in；When the application record data that are completed do not include the alternative unfinished application record data
In User Identity, then judge no matching result；Alternatively, when the application record data that are completed include described alternative
Unfinished application record data in User Identity, continue judgement described in be completed whether application record data include institute
The crucial request for data in alternative unfinished application record data is stated, the key request for data is for distinguishing different Shens
Please；If the application record data that are completed are not comprising the crucial application number in the alternative unfinished application record data
According to then judging no matching result.Wherein, User Identity may include user identity card number, passport No., cell-phone number, postal
Case, social application account or IMEI number etc..Crucial request for data can distinguish different financial products, such as name of product, production
Product code name etc..
Furthermore it is also possible to alternative unfinished application record data are matched based on application record data are completed,
The entry that application record data are completed in matching in the alternative unfinished application record data is weeded out, to filter out not
Complete application record data.
Specifically, it can first obtain in unfinished application record data alternative in first database and identify unfinished Shen
It please record, by the unfinished application record, (including registration information and other information, the same person may apply for multiple gold
Melt product, to be distinguish) (the step of duplicate removal being increased) is matched in the second database, pass through User Identity
It searches, does not include use in the alternative unfinished application record data when application record data are completed in the second database
Family identity then judges no matching result；Alternatively, when the application record data that be completed include described alternative not complete
At the User Identity in application record data, continue to be completed whether application record data include described alternative described in judgement
Unfinished application record data in crucial request for data, the key request for data is for distinguishing different applications；If
The application record data that are completed then are sentenced not comprising the crucial request for data in the alternative unfinished application record data
Break no matching result.
The unfinished application record data input user portrait label model is carried out operation, with output by step S130
User's portrait label of the corresponding unfinished application record data.
Wherein, user's portrait label model is that the pre- mode for first passing through training data training machine learning model obtains
It arrives.The purpose of user tag is the type in order to describe user, and the purpose of user's portrait is the demand in order to predict user, because
This user label of drawing a portrait is to need sum-type to embody prediction and describe user, and illustrating, it is specifically to predict user letter of application
Loan demand and/or user type.
In particular it is required that the mode training one for first passing through machine learning in advance is used to generate the model of user's portrait label.
When the application record data that input one does not complete, the exportable corresponding user's portrait label of the model.The process of model training
It needs that data are first arranged in client-side program and buries a little, or setting data bury mode a little in the application page, acquire each user
Behavioral data in application process is then based on each characteristic dimension needs of training data, to user's application process of acquisition
In behavioral data be labeled, to generate a large amount of training data for not completing application user；To above-mentioned a large amount of training
Data carry out machine learning model training, finally obtain user's portrait label model.Wherein, the user draws a portrait label including such as
Under one or more combination: trial type client, privacy-sensitive type client lack credit data type client, potential deceptive information
Client, and/or it can not judge type customer etc..
Accordingly, behavior of these the above-mentioned different types of clients before abandoning registration has certain feature, such as tastes
Examination type client be mostly just abandon having registered when registering essential information, this is because application loan need the essential information filled in compared with
More, the client (i.e. user) of non-genuine demand tends not to adhere to completing registration information fill flow path；For another example privacy-sensitive type
Client may abandon providing a loan due to being reluctant to input whole individual privacy informations, and person is by burying such client of point data tracking display
It can be abandoned in the stage of input individual privacy information；Note that not making the above is only example client portrait and potential registration behavior
For limitation of the invention.
Further, different customer types can be defined, and registration behavior is also not by way of manually setting,
But by a large amount of application user data that do not complete of mark as training data, and pass through machine learning mode one user of training
Label model is drawn a portrait to complete.
User's portrait label model can be based on two classification, decision tree, cluster, Bayes's classification, supporting vector
Machine SVM, EM, Adaboost and neural network etc. one or more combinations is realized, the unfinished Shen of input is analyzed
The behavioral data in each user's application process for including in data please be record, based in the behavioral data in each user's application process
One or more behavioural characteristics, sort operation is carried out to the unfinished application record data of each user of input, is corresponded to
The feature tag of application record data is not completed.
The corresponding label of unfinished application record data of each user is obtained after the sort operation, i.e., draws a portrait as user
Label exports user's portrait label corresponding with each record user identifier of data.
Wherein, user's portrait label includes following one or more group based on machine learning model operation output
Close: trial type client, privacy-sensitive type client lack credit data type client, potential deceptive information client, and/or can not sentence
Disconnected type customer etc..
Step S140 draws a portrait label to user belonging to the unfinished application record data according to the user, pushes
Customized information.
Wherein, to various types of users portrait label of model output, the user that mark is predicted as the same user is drawn
Picture.Among these using User ID as user's unique identification, summarizes the unfinished record data of unification user ID, pass through the user
It draws a portrait after model, counts all user's portrait labels relevant to same User ID, it is same by these relevant label for labelling
The user of user draws a portrait, so that it is determined that user type belonging to the user.
According to user type belonging to the user, the customized information of type matching is obtained in customized information library.Further,
After obtaining user type, system will send corresponding customized information according to different type, concrete ways may include short message,
Wechat, mail etc. get corresponding cell-phone number, WeChat ID here by user in the essential information that the application stage fills in
And email address, the terminal device that matched customized information pushing is used to the user.For example, for trial type client,
System will send common favor information；For privacy-sensitive type client, often customer value is higher, and customized information is more
It provides without guarantee, the financial product without private data；By above method, system will be unfinished in first database
The user of application automatically analyzes, and distributes corresponding customized information, accelerates to convert the user that this part can not promote originally.
Referring to FIG. 2, one embodiment that the present invention passes through customized information pushing system, comprising:
Alternative recording unit 210, for choosing alternative unfinished application record data；
Data matching unit 220, for by the alternative unfinished application record data and application record number being completed
According to being matched；
Determination unit 230, for when not having matching result, the alternative unfinished application record data to be determined as
Application record data are not completed
Model arithmetic unit 240, for carrying out the unfinished application record data input user portrait label model
Operation, to export user's portrait label of the corresponding unfinished application record data；
Push unit 250, for being remembered to the unfinished application according to the relevant user type of user portrait label
It records user belonging to data and pushes customized information.
Specifically, the customized information pushing system, further includes: first database is buried for storing by the way that data are arranged
Related data in user's application process of point acquisition；The alternative recording unit is further used for the institute from first database
It states and chooses alternative unfinished application record data in related data.For data are set in client-side program and are buried a little, in advance
First it is related to business framework setting data on user's application process path in client development process to bury a little, obtains user Shen
It please the tracking number of total interface parameter and/or method parameter that is related to of the client traffic call chain road that passes through of service traffics
According to acquiring out the behavioral data in each user's application process based on above-mentioned tracking data.Here behavioral data can be according to tool
Body needs to be arranged different types of feature, such as when the residence time in essential information page, project input sequence, click next step
Check do not fill out item, identification card number format checking, cell-phone number format check, mailbox format check, Bank Account Number format check,
The page, the verification of proof attachment and/or request times etc. where individual privacy information page residence time, abandon application registration.It is above-mentioned
These behavioral datas can be got by way of data are arranged in client-side program and bury a little, similarly, compared to visitor
Family end program, only application the page in setting data bury mode a little possibility limitation it is higher, the behavioral data type got
Feature is less, but nor affects on and realize method of the invention.
Further, in data structure, setting includes the second database that completed application record is collected in storage, with
And the first database of the unfinished application record of storage alternatively.In general, the second database is when user applies completing
By the database of record deposit, for example registration center is set or applies for the database at center, data content value is larger
Also more important, therefore its second database and other databases are in the state being relatively isolated.And first database records
Be the i.e. alternative unfinished application process of all users behavioral data, i.e., either do not complete application Shen is still completed
Will collect and record please.For example, unfinished application record can be stamped into " unfinished " label in first database,
But this label may be inaccuracy, therefore the completed application record recorded in the second database is by other approach
Verification was carried out, this process is different from the data of first database, and even the second database is to pass through separate sources in other words
Data included.It is tested in this way, the data in first database need to carry out matching according to the data in the second database
Card, and further screening is done,
Specifically, the matching unit, is also used to, and is completed whether application record data include described alternative described in judgement
Unfinished application record data in User Identity；When it is described be completed application record data do not include it is described alternative
The User Identity in application record data is not completed, then judges no matching result；Alternatively, application note is completed when described
Recording data includes the User Identity in the alternative unfinished application record data, continues that application is completed described in judgement
Whether record data include crucial request for data in the alternative unfinished application record data, the key request for data
For distinguishing different applications；If the application record data that are completed do not include the alternative unfinished application record number
Crucial request for data in, then judge no matching result.Wherein, User Identity may include user identity card number,
Passport No., cell-phone number, mailbox, social application account or IMEI number etc..Crucial request for data can distinguish different financial products,
Such as name of product, product designation etc..
Furthermore it is also possible to based on application record data are completed to the alternative unfinished application record data progress
Match, the entry that application record data are completed in matching in the alternative unfinished application record data is weeded out, to screen
Application record data are not completed out.
Specifically, it can first obtain in unfinished application record data alternative in first database and identify unfinished Shen
It please record, by the unfinished application record, (including registration information and other information, the same person may apply for multiple gold
Melt product, to be distinguish) (the step of duplicate removal being increased) is matched in the second database, pass through User Identity
It searches, does not include use in the alternative unfinished application record data when application record data are completed in the second database
Family identity then judges no matching result；Alternatively, when the application record data that be completed include described alternative not complete
At the User Identity in application record data, continue to be completed whether application record data include described alternative described in judgement
Unfinished application record data in crucial request for data, the key request for data is for distinguishing different applications；If
The application record data that are completed then are sentenced not comprising the crucial request for data in the alternative unfinished application record data
Break no matching result.
Further, the customized information pushing system, further includes: acquisition unit, for by client-side program
Setting data are buried a little, or setting data bury mode a little in the application page, acquire the behavioral data in each user's application process；
Training data generation unit is labeled for the behavioral data in user's application process to acquisition, is generated a large amount of not complete
At the training data of application user；Model training unit is obtained for carrying out machine learning model training using the training data
To user's portrait label model.
Wherein, user's portrait label model is that the pre- mode for first passing through training data training machine learning model obtains
It arrives.The purpose of user tag is the type in order to describe user, and the purpose of user's portrait is the demand in order to predict user, because
This user label of drawing a portrait is to need sum-type to embody prediction and describe user, and illustrating, it is specifically to predict user letter of application
Loan demand and/or user type.
In particular it is required that the mode training one for first passing through machine learning in advance is used to generate the model of user's portrait label.
When the application record data that input one does not complete, the exportable corresponding user's portrait label of the model.The process of model training
It needs that data are first arranged in client-side program and buries a little, or setting data bury mode a little in the application page, acquire each user
Behavioral data in application process is then based on each characteristic dimension needs of training data, to user's application process of acquisition
In behavioral data be labeled, to generate a large amount of training data for not completing application user；To above-mentioned a large amount of training
Data carry out machine learning model training, finally obtain user's portrait label model.Wherein, the user draws a portrait label including such as
Under one or more combination: trial type client, privacy-sensitive type client lack credit data type client, potential deceptive information
Client, and/or it can not judge type customer etc..
Accordingly, behavior of these the above-mentioned different types of clients before abandoning registration has certain feature, such as tastes
Examination type client be mostly just abandon having registered when registering essential information, this is because application loan need the essential information filled in compared with
More, the client (i.e. user) of non-genuine demand tends not to adhere to completing registration information fill flow path；For another example privacy-sensitive type
Client may abandon providing a loan due to being reluctant to input whole individual privacy informations, and person is by burying such client of point data tracking display
It can be abandoned in the stage of input individual privacy information；Note that not making the above is only example client portrait and potential registration behavior
For limitation of the invention.
Further, different customer types can be defined, and registration behavior is also not by way of manually setting,
But by a large amount of application user data that do not complete of mark as training data, and pass through machine learning mode one user of training
Label model is drawn a portrait to complete.
User's portrait label model can be based on two classification, decision tree, cluster, Bayes's classification, supporting vector
Machine SVM, EM, Adaboost and neural network etc. one or more combinations is realized, the unfinished Shen of input is analyzed
The behavioral data in each user's application process for including in data please be record, based in the behavioral data in each user's application process
One or more behavioural characteristics, sort operation is carried out to the unfinished application record data of each user of input, is corresponded to
The feature tag of application record data is not completed.
Wherein, the model arithmetic unit, is also used to, and the unfinished application record of each user is obtained after the sort operation
The corresponding user's portrait label of data, exports user's portrait label corresponding with each record user identifier of data.The classification
The corresponding label of unfinished application record data of each user is obtained after operation, i.e., as user's portrait label, output and each note
Record the corresponding user's portrait label of user identifier of data.Further, the user draws a portrait label based on machine learning model
Operation output includes following one or more combination: trial type client, privacy-sensitive type client lack credit data type visitor
Family, potential deceptive information client, and/or it can not judge type customer etc..
Wherein, the push unit, is also used to, using User ID as user's unique identification, statistics and same User ID phase
User's portrait label of pass determines user type belonging to the user by respective labels mark user's portrait.Further, right
Various types of users portrait label of model output, mark are predicted as user's portrait of the same user.Among these with user
ID summarizes the unfinished record data of unification user ID as user's unique identification, after user portrait model, statistics
All user's portrait labels relevant to same User ID, are drawn a portrait by the user of these relevant same users of label for labelling,
So that it is determined that user type belonging to the user.
In addition, the user type according to belonging to the user, obtains the customized information of type matching in customized information library.Into
One step, after obtaining user type, system will send corresponding customized information according to different types, and concrete ways can wrap
Short message, wechat, mail etc. are included, gets corresponding mobile phone in the essential information that the application stage fills in here by user
Number, WeChat ID and email address, the terminal device that matched customized information pushing is used to the user.For example, for tasting
Examination type client, system will send common favor information；For privacy-sensitive type client, often customer value is higher, customization letter
Breath is more to provide without guarantee, the financial product without private data；By above method, system will be in first database
In the user of unfinished application automatically analyze, and distribute corresponding customized information, accelerating to convert this part originally can not
The user of popularization.
Another embodiment of the present invention additionally provides a kind of Fig. 3 and shows a kind of meter provided according to an embodiment of the present invention
Calculate device structure schematic diagram.The calculating equipment 300 includes: processor 310, and is stored with and can transport on the processor 310
The memory 320 of capable computer program.Processor 310, for being held when executing the computer program in the memory 320
Each step of method in the row present invention.Memory 320 can be (the read-only storage of electrically erasable of such as flash memory, EEPROM
Device), the electronic memory of EPROM, hard disk or ROM etc.Memory 320 has storage for executing appointing in the above method
The memory space 330 of the computer program 331 of what method and step.Computer program 331 can be from one or more computer
It reads or is written in program product in this one or more computer program product.These computer program products include
Such as hard disk, the program code carrier of compact-disc (CD), storage card or floppy disk etc.For ease of description, illustrate only with
The relevant part of this specification embodiment, it is disclosed by specific technical details, please refer to this specification embodiment method part.It should
Equipment is calculated, can be the calculating equipment formed including various electronic equipments, PC computer, network Cloud Server or even mobile phone are put down
(Point of Sales, sale is eventually by plate computer, PDA (Personal Digital Assistant, personal digital assistant), POS
End), vehicle-mounted computer, the server capability being arranged on any electronic equipment such as desktop computer.
Based on this understanding, this specification realizes all or part of the process in the method for above-mentioned first embodiment,
Relevant hardware can be instructed to complete by computer program, it is computer-readable that the computer program can be stored in one
In storage medium, the computer program is when being executed by processor, it can be achieved that the step of above-mentioned each embodiment of the method.Wherein,
The computer program includes computer program code, and the computer program code can be source code form, object identification code
Form, executable file or certain intermediate forms etc..The computer-readable medium may include: that can carry the computer
Any entity or device of program code, medium, USB flash disk, mobile hard disk, magnetic disk, CD, computer storage, read-only memory
(ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, electricity
Believe signal and software distribution medium etc..It should be noted that the content that the computer-readable medium includes can be according to department
Make laws in method administrative area and the requirement of patent practice carry out increase and decrease appropriate, such as in certain jurisdictions, according to legislation and
Patent practice, computer-readable medium do not include electric carrier signal and telecommunication signal.
Although the preferred embodiment of this specification has been described, once a person skilled in the art knows basic wounds
The property made concept, then additional changes and modifications may be made to these embodiments.So the following claims are intended to be interpreted as includes
Preferred embodiment and all change and modification for falling into this specification range.
Obviously, those skilled in the art can carry out various modification and variations without departing from this specification to this specification
Spirit and scope.In this way, if these modifications and variations of this specification belong to this specification claim and its equivalent skill
Within the scope of art, then this specification is also intended to include these modifications and variations.
Claims (16)
1. a kind of customized information pushing method characterized by comprising
Choose alternative unfinished application record data；
The alternative unfinished application record data are matched with application record data are completed；
When there is no matching result, the alternative unfinished application record data are determined as not completing application record data；
The unfinished application record data input user portrait label model is subjected to operation, described do not complete is corresponded to output
User's portrait label of application record data；
Customized information is pushed to user belonging to the unfinished application record data according to user portrait label.
2. the method as described in claim 1, which is characterized in that described to choose alternative unfinished application record data, comprising:
Alternative unfinished application record data are chosen from first database, the data of the first database storage include logical
It crosses setting data and buries related data in user's application process of acquisition.
3. method according to claim 1 or 2, which is characterized in that described by the alternative unfinished application record data
It is matched with application record data are completed, further comprises:
It is completed whether application record data include user's body in the alternative unfinished application record data described in judgement
Part mark；
When the application record data that are completed are not comprising the user identity mark in the alternative unfinished application record data
Know, then judges no matching result.
4. method according to claim 1 or 2, which is characterized in that described by the alternative unfinished application record data
It is matched with application record data are completed, further comprises:
When the application record data that are completed include the User Identity in the alternative unfinished application record data,
Continue to be completed whether application record data include crucial Shen in the alternative unfinished application record data described in judgement
Please data, the key request for data is for distinguishing different applications；
If the application record data that are completed are not comprising the crucial application in the alternative unfinished application record data
Data then judge no matching result.
5. the method as described in claim 1, which is characterized in that user label model of drawing a portrait is pre- to first pass through training data
The mode of machine learning model training obtains, and further comprises:
Related data in user's application process needed for burying an acquisition training pattern by setting data；
Related data in user's application process of acquisition is labeled, a large amount of training number for not completing application user is generated
According to；
Machine learning model training is carried out using the training data, obtains user's portrait label model.
6. the method as described in claim 1, which is characterized in that user's portrait label includes following one or more group
Close: trial type client, privacy-sensitive type client lack credit data type client, potential deceptive information client, and/or can not sentence
Disconnected type customer.
7. the method as described in claim 1, which is characterized in that the unfinished application record data input user draws a portrait and marks
It signs model and carries out operation, further comprise:
Each user for including in the unfinished application record data of user's portrait label model analysis input applied
Behavioral data in journey, based on one or more behavioural characteristics in the behavioral data in each user's application process, to input
The unfinished application record data of each user carry out sort operation.
8. a kind of customized information pushing system characterized by comprising
Alternative recording unit, for choosing alternative unfinished application record data；
A data matching unit, for by the alternative unfinished application record data and application record data being completed carrying out
Match；
Determination unit, for when not having matching result, the alternative unfinished application record data being determined as not completing
Application record data；
Model arithmetic unit, for the unfinished application record data input user portrait label model to be carried out operation, with
User's portrait label of the corresponding unfinished application record data of output；
Push unit is used for the relevant user type of label of drawing a portrait according to the user, to the unfinished application record data
Affiliated user pushes customized information.
9. system as claimed in claim 8, which is characterized in that further include:
First database, for storing the related data buried in user's application process of acquisition by setting data；
The alternative recording unit is further used for choosing alternative unfinished Shen from related data described in first database
It please record data.
10. system as claimed in claim 8, which is characterized in that the matching unit is also used to,
It is completed whether application record data include user's body in the alternative unfinished application record data described in judgement
Part mark；When the application record data that are completed are not comprising the user identity in the alternative unfinished application record data
Mark, then judge no matching result.
11. system as claimed in claim 8, which is characterized in that the matching unit is also used to,
When the application record data that are completed include the User Identity in the alternative unfinished application record data,
Continue to be completed whether application record data include crucial Shen in the alternative unfinished application record data described in judgement
Please data, the key request for data is for distinguishing different applications；
If the application record data that are completed are not comprising the crucial application in the alternative unfinished application record data
Data then judge no matching result.
12. system as claimed in claim 8, which is characterized in that user label model of drawing a portrait is pre- to first pass through trained number
It is obtained according to the mode of training machine learning model；
The system, further includes:
Acquisition unit, for by client-side program be arranged data bury a little, or application the page in setting data bury a little
Mode acquires the behavioral data in each user's application process；
Training data generation unit is labeled for the behavioral data in user's application process to acquisition, is generated a large amount of
The training data of application user is not completed；
Model training unit obtains user's portrait label mould for carrying out machine learning model training using the training data
Type.
13. the system as described in right wants 8, which is characterized in that user's portrait label includes following one or more group
Close: trial type client, privacy-sensitive type client lack credit data type client, potential deceptive information client, and/or can not sentence
Disconnected type customer.
14. system as claimed in claim 8, which is characterized in that the model arithmetic unit is also used to,
Each user for including in the unfinished application record data of user's portrait label model analysis input applied
Behavioral data in journey, based on one or more behavioural characteristics in the behavioral data in each user's application process, to input
The unfinished application record data of each user carry out sort operation.
15. a kind of server, including processor and memory:
The memory is used to store the program that perform claim requires any one of 1 to 7 the method；
The processor is configured to for executing the program stored in the memory.
16. a kind of computer readable storage medium, is stored with computer program, which is characterized in that the program is held by processor
The step of any one of claim 1 to 7 the method is realized when row.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910017478.8A CN109783731B (en) | 2019-01-08 | 2019-01-08 | Custom information pushing method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910017478.8A CN109783731B (en) | 2019-01-08 | 2019-01-08 | Custom information pushing method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109783731A true CN109783731A (en) | 2019-05-21 |
| CN109783731B CN109783731B (en) | 2024-05-17 |
Family
ID=66500228
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910017478.8A Active CN109783731B (en) | 2019-01-08 | 2019-01-08 | Custom information pushing method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109783731B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112149029A (en) * | 2019-06-26 | 2020-12-29 | 北京百度网讯科技有限公司 | Method and device for determining user value viewing label |
| CN112440566A (en) * | 2019-08-28 | 2021-03-05 | 精工爱普生株式会社 | Recording condition determining method and recording apparatus |
| CN112927059A (en) * | 2021-04-21 | 2021-06-08 | 邓文浩 | Personalized digital type zero work service recommendation method and system based on deep learning |
| CN116385052A (en) * | 2023-02-07 | 2023-07-04 | 网舟联合科技(北京)有限公司 | Method for analyzing potential user characteristics based on Bayesian prediction embedded point document value |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102595306A (en) * | 2012-02-10 | 2012-07-18 | 北京无限新锐网络科技有限公司 | System and method for providing personalized health messages |
| CN103337027A (en) * | 2013-06-20 | 2013-10-02 | 咸晓梦 | Clothing information pushing method and system |
| CN104735062A (en) * | 2015-03-12 | 2015-06-24 | 微梦创科网络科技（中国）有限公司 | Network user registration method and server |
| CN104967552A (en) * | 2014-11-12 | 2015-10-07 | 腾讯科技（深圳）有限公司 | Message push method and apparatus |
| CN107092452A (en) * | 2016-02-17 | 2017-08-25 | 阿里巴巴集团控股有限公司 | A kind of information processing method and device of displaying business completion status |
| CN107291728A (en) * | 2016-03-31 | 2017-10-24 | 北京金山安全软件有限公司 | Method and device for acquiring user attribute and electronic equipment |
-
2019
- 2019-01-08 CN CN201910017478.8A patent/CN109783731B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102595306A (en) * | 2012-02-10 | 2012-07-18 | 北京无限新锐网络科技有限公司 | System and method for providing personalized health messages |
| CN103337027A (en) * | 2013-06-20 | 2013-10-02 | 咸晓梦 | Clothing information pushing method and system |
| CN104967552A (en) * | 2014-11-12 | 2015-10-07 | 腾讯科技（深圳）有限公司 | Message push method and apparatus |
| CN104735062A (en) * | 2015-03-12 | 2015-06-24 | 微梦创科网络科技（中国）有限公司 | Network user registration method and server |
| CN107092452A (en) * | 2016-02-17 | 2017-08-25 | 阿里巴巴集团控股有限公司 | A kind of information processing method and device of displaying business completion status |
| CN107291728A (en) * | 2016-03-31 | 2017-10-24 | 北京金山安全软件有限公司 | Method and device for acquiring user attribute and electronic equipment |
Non-Patent Citations (1)
| Title |
|---|
| 贾伟洋: ""基于群组用户画像的农业信息化推荐算法研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112149029A (en) * | 2019-06-26 | 2020-12-29 | 北京百度网讯科技有限公司 | Method and device for determining user value viewing label |
| CN112149029B (en) * | 2019-06-26 | 2023-07-07 | 北京百度网讯科技有限公司 | Method and device for determining user value viewing labels |
| CN112440566A (en) * | 2019-08-28 | 2021-03-05 | 精工爱普生株式会社 | Recording condition determining method and recording apparatus |
| CN112440566B (en) * | 2019-08-28 | 2023-07-07 | 精工爱普生株式会社 | Recording condition determining method and recording device |
| CN112927059A (en) * | 2021-04-21 | 2021-06-08 | 邓文浩 | Personalized digital type zero work service recommendation method and system based on deep learning |
| CN116385052A (en) * | 2023-02-07 | 2023-07-04 | 网舟联合科技(北京)有限公司 | Method for analyzing potential user characteristics based on Bayesian prediction embedded point document value |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109783731B (en) | 2024-05-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109783731A (en) | A kind of customized information pushing method and system | |
| CN111915366B (en) | User portrait construction method, device, computer equipment and storage medium | |
| CN109493199A (en) | Products Show method, apparatus, computer equipment and storage medium | |
| US20160379268A1 (en) | User behavior data analysis method and device | |
| CN111178219A (en) | Bill identification management method and device, storage medium and electronic equipment | |
| CN111091351A (en) | User portrait construction method and device, electronic equipment and readable storage medium | |
| CN110798567A (en) | Short message classification display method and device, storage medium and electronic equipment | |
| CN109800354B (en) | Resume modification intention identification method and system based on block chain storage | |
| CN107480858A (en) | A kind of Aided intelligent decision-making and method based on the analysis of stock big data | |
| CN110033120A (en) | For providing the method and device that risk profile energizes service for trade company | |
| CN107742128A (en) | Method and apparatus for output information | |
| CN112861980B (en) | Calendar task table mining method based on big data and computer equipment | |
| CN112989188A (en) | Determination method and device of recommended order and server | |
| WO2017049917A1 (en) | Email processing method and apparatus | |
| CN115238688A (en) | Electronic information data association relation analysis method, device, equipment and storage medium | |
| CN112163155A (en) | Information processing method, device, equipment and storage medium | |
| CN113327132A (en) | Multimedia recommendation method, device, equipment and storage medium | |
| US11561963B1 (en) | Method and system for using time-location transaction signatures to enrich user profiles | |
| CN116523069A (en) | Coupon pushing method and device based on federal modeling | |
| CN115660733A (en) | Sales prediction system and method based on artificial intelligence | |
| CN114066513A (en) | User classification method and device | |
| CN111382343B (en) | Label system generation method and device | |
| CN113688190A (en) | Method and device for acquiring customer relationship | |
| CN114510638B (en) | Information processing method, apparatus, device, storage medium, and program product | |
| CN109886616A (en) | Data processing method, device, equipment and storage medium between multisystem |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |