CN109325845A - A kind of financial product intelligent recommendation method and system - Google Patents
A kind of financial product intelligent recommendation method and system Download PDFInfo
- Publication number
- CN109325845A CN109325845A CN201810939473.6A CN201810939473A CN109325845A CN 109325845 A CN109325845 A CN 109325845A CN 201810939473 A CN201810939473 A CN 201810939473A CN 109325845 A CN109325845 A CN 109325845A
- Authority
- CN
- China
- Prior art keywords
- user
- group
- different
- behavioural characteristic
- preference
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/02—Banking, e.g. interest calculation or account maintenance
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present embodiments relate to a kind of intelligent recommendation methods, comprising: acquisition user behavior data and collage-credit data；Generate behavioural characteristic label and reference feature tag respectively for user；User is divided by different user behavior groups and different user's reference groups according to behavioural characteristic label and reference feature tag respectively；The preference of different user behavior group is calculated according to behavioural characteristic label；The air control rule match degree of different user reference group is calculated according to reference feature tag；Financial product recommendation is carried out to different users in conjunction with the air control rule match degree of user behavior group preference and user's reference group；Existing behavioral data difference regeneration behavior feature tag and reference feature tag based on the user, while user's owning user group is redistributed, and calculate the preference and air control rule match degree of groups of users after update, Automatic Optimal recommended models.Further to a kind of intelligent recommendation system.
Description
Technical field
The invention belongs to data processing field more particularly to a kind of financial product intelligent recommendation method and system.
Background technique
With the fast development of Internet technology, personalized recommendation has become net indispensable in people's network life
One of network service also becomes the emphasis of internet product future development.
Currently, the recommended method for product mainly includes two kinds: the first, user is according to own interests setting hobby
Interest tags, system are that user recommends according to the matching degree of recommendation information and user interest label.Second is user
According to own interests, interested class label is selected in the classification set, system is according to recommendation information and user
The matching degree of interest tags is that user recommends.
But above two recommended method all has the following problems: since user is unwilling to take more time and energy
It goes to be set or even some users is to choose at random label, make label description inaccuracy, can not recommend more accurately for user
Requirement product, so as to cause the reduction of recommendation results accuracy rate.
In view of this, nowadays there is an urgent need to design a kind of intelligent recommendation method, to solve to recommend accuracy in the prior art
Lower low shortcoming.
Summary of the invention
The embodiment of the invention provides a kind of financial product intelligent recommendation method and system, to solve to exist in the prior art
The lower problem of recommendation accuracy rate.
The embodiment of the present invention provides a kind of financial product intelligent recommendation method comprising following steps:
Acquire the behavioral data and collage-credit data of user；
Based on the behavioral data and collage-credit data, behavioural characteristic label and reference feature mark are generated respectively for the user
Label；
User is divided into according to the behavioural characteristic label and reference feature tag by different user behavior groups respectively
And different user's reference group；
Different user behavior group, which is calculated, according to the behavioural characteristic label of different user behavior group corresponds to each gold
Melt the preference of product；
Different user reference group, which is calculated, according to the reference feature tag of different user reference group corresponds to each gold
Melt the air control rule match degree of product；
In conjunction with the air control rule match degree of user behavior group preference and user's reference group to different
User carries out financial product recommendation；
Existing behavioral data based on the user updates the behavioural characteristic label and reference feature mark of the user respectively
Label, while user's owning user group is redistributed, and calculate the preference and air control rule of groups of users after update
With degree, Automatic Optimal recommended models.
Further, the behavioral data includes the location information and/or application installation preference information of user.
Further, the collage-credit data includes personal reference report, public accumulation fund data and social security card data.
Further, when generating the behavioural characteristic label or reference feature tag, to the behavioral data or reference
Data carry out word segmentation processing to extract Feature Words.
Further, different user behavior group is calculated according to the behavioural characteristic label of different user behavior group to correspond to respectively
The method of the preference of a financial product specifically includes:
Determine the weight of the different behavioural characteristic labels of same user behavior group；
Different behavioural characteristic label weights is sorted from large to small；
Using the weight limit of behavioural characteristic label as the user behavior group to the preference of financial product.
Further, described that different user group pair is calculated according to the reference feature tag of different user reference group
The method for answering the air control rule match degree of each financial product specifically includes:
Based on the reference feature tag, different credit angle value is assigned to different user reference group；
The credit angle value is ranked up；
According to the risk class for determining different user reference group that sorts.
Further, the existing behavioral data of the user include user investment behavior data and borrow refund behavior number
According to.
In addition, the embodiment of the present invention provides a kind of intelligent recommendation system of financial product simultaneously, comprising:
Acquisition module, for acquiring the behavioral data and collage-credit data of user；
Generation module generates behavioural characteristic mark for being based on the behavioral data and collage-credit data for the user respectively
Label and reference feature tag；
User is divided into for respectively according to the behavioural characteristic label and reference feature tag different by division module
User behavior group and different user's reference groups；
Preference computing module, for calculating different user according to the behavioural characteristic label of different user behavior group
Behavior group corresponds to the preference of each financial product；
Air control matching degree computing module, it is different for being calculated according to the reference feature tag of different user reference group
User's reference group corresponds to the air control rule match degree of each financial product；
Recommending module, for the air control rule in conjunction with user behavior group preference and user's reference group
Financial product recommendation is carried out to different users with degree；
Update module updates the behavioural characteristic label of the user for the existing behavioral data based on the user respectively
With reference feature tag, while redistributing user's owning user group, and calculate update after groups of users preference
With air control rule match degree, Automatic Optimal recommended models.
Further, the behavioral data includes the location information and/or application installation preference information of user.
Further, the existing behavioral data of the user include user investment behavior data and borrow refund behavior number
According to.
The above scheme of the embodiment of the present invention compared with prior art, has the advantages that the embodiment of the present invention mentions
The financial product intelligent recommendation method of confession by generating behavioural characteristic label and reference feature tag according to user information, and is divided
User is not grouped according to the behavioural characteristic label and reference feature tag, calculate separately different groups preference and
Air control rule match degree；Products Show is carried out in conjunction with the preference of different groups and the air control rule match degree of different groups；Afterwards
Continuous real-time tracing and update, and Automatic Optimal recommended models.This method is marketed different products for different target user, is improved
The accuracy of recommendation results, so as to more accurately recommend the personalized product for being more bonded user itself for user；Separately
It also can be improved the marketing efficiency of product outside.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly introduced, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this
For the those of ordinary skill in field, without any creative labor, it can also be obtained according to these attached drawings
His attached drawing.
Fig. 1 is the flow chart of financial product intelligent recommendation method of the embodiment of the present invention.
Fig. 2 is to calculate different user behavior according to the behavioural characteristic label of different user behavior group in the embodiment of the present invention
Group corresponds to the method flow diagram of the preference of each financial product.
Fig. 3 is to calculate different user according to the reference feature tag of different user reference group in the embodiment of the present invention
Group corresponds to the method flow diagram of the air control rule match degree of each financial product.
Fig. 4 is the structural schematic diagram of financial product intelligent recommendation system of the embodiment of the present invention.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with attached drawing to the present invention make into
It is described in detail to one step, it is clear that described embodiments are only a part of the embodiments of the present invention, rather than whole implementation
Example.Based on the embodiments of the present invention, obtained by those of ordinary skill in the art without making creative efforts
All other embodiment, shall fall within the protection scope of the present invention.
The preferred embodiment that the invention will now be described in detail with reference to the accompanying drawings.
Embodiment 1
Referring to Fig. 1, being a kind of flow chart of financial product intelligent recommendation method provided in an embodiment of the present invention, this method
Include:
S101 acquires the behavioral data and collage-credit data of user.
Wherein, the premise of the acquisition various data of user is must sufficiently to authorize through user.After user sufficiently authorizes,
Necessary user behavior data and collage-credit data can be acquired.
In the embodiment of the present invention, the behavioral data includes but is not limited to the location information and/or application installation of user
Preference information etc..The location information of the user can obtain the location information of mobile terminal user based on location-based service (LBS).Institute
State using installation preference information include in certain time inner electronic equipment the installation number of application program and unloading quantity, each
Installation number and unloading number of application program etc..The behavioral data of the user can also include that the click of user browses number
According to, browsing duration data, browsing range data etc..
The collage-credit data includes personal reference report, public accumulation fund data and social security card data etc..The collage-credit data can
It is acquired by Third Party Authentication.The third party can have for Central Bank's reference center, public accumalation fund for housing construction center, social security office etc. to be recognized
Demonstrate,prove the mechanism of qualification.
S102 is based on the behavioral data and collage-credit data, generates behavioural characteristic label and reference respectively for the user
Feature tag.
Wherein, after the behavioral data and collage-credit data for collecting user, the behavioral data according to the user is system
The user generates behavioural characteristic label, while being that the user generates reference feature mark according to the collage-credit data of the user
Label.Generate label for different user, facilitate it is subsequent user is sorted out, to carry out optimal push away to different user group
It recommends.
When generating the behavioural characteristic label or reference feature tag, can to the behavioral data or collage-credit data into
Row word segmentation processing is to extract Feature Words.Carrying out word segmentation processing to the behavioral data or collage-credit data can be public using this field
The participle technique known carries out, and the Feature Words extracted are exactly the word obtained after word segmentation processing, or is known as key
Word.
In the embodiment of the present invention, for the behavioural characteristic label, each user has at least one behavioural characteristic mark
Label, behavior feature tag is the keyword that can summarize the main contents of behavioral data of the user.For example, false
If the behavioral data shows " location information of the user is Changsha ", " star Zhang Jie July 13 open concert in Changsha " etc.,
These data contents analyze and produce the feature tags such as " Changsha ", " amusement "；Assuming that described in the behavioral data shows
User is mounted with the application programs such as " grass roots investment ", " 360 wealth ", " realizing doctrinairism wealth ", " where go ", " taking journey " recently, to this
A little data, which analyze, produces the feature tags such as " financing ", " tourism ".
For the reference feature tag, which is the pass that can summarize the credit rating of the user
Keyword.It illustrating, it is assumed that it is higher that individual subscriber reference report is displayed without overdue record, five one gold medals of danger pay number,
Then generate feature tags such as " top-tier customers ".
User is divided into different user behaviors according to the behavioural characteristic label and reference feature tag respectively by S103
Group and different user's reference groups；
It wherein, can be by user according to the label after the behavioural characteristic label and reference feature tag for generating user
Sorted out.In the embodiment of the present invention, corresponding user is on the one hand divided by different groups according to the behavioural characteristic label,
That is user behavior group；On the other hand, corresponding user is divided by different groups, i.e. user according to the reference feature tag
Reference group.
Preferably, can will whether similar or close according to the keyword of the feature tag, will be similar or similar
User is classified as one kind.For example, being " Changsha ", the user of " amusement " or with similar tags by the user behavior characteristics label
User can be classified as a groups of users, be " financing " by behavioural characteristic label, " tourism " or user with similar tags can
It is classified as a groups of users.The user that user's reference feature tag is " top-tier customer " can be classified as a groups of users,
The user that reference feature tag is " medium client " is classified as a groups of users.
It is corresponding each to calculate different user behavior group according to the behavioural characteristic label of different user behavior group by S104
The preference of a financial product；
Wherein, the financial product can be understood as including different finance products or different loan products.The use
The preference of family behavior group can be understood as the user in the different user group to the degree of concern of financial product.To certain
A finance product degree of concern is higher, and the customer investment purchase intention represented in the groups of users is very high；Degree of concern is not high,
Represent user in the groups of users not no intention of Investment & Financing recently.
In the embodiment of the present invention, as shown in Fig. 2, calculating different use according to the behavioural characteristic label of different user behavior group
The method for the preference that family behavior group corresponds to each financial product specifically includes step:
S201 determines the weight of the different behavioural characteristic labels of same user behavior group.It wherein, can be according to the difference
The probability that the Feature Words of behavioural characteristic label occur carries out the assignment of weight.
S202 sorts from large to small different behavioural characteristic label weights；
S203, using the weight limit of behavioural characteristic label as the user behavior group to the preference of financial product.
For example, if the feature tag of a groups of users includes " loan ", " being located at Beijing " and " travel abroad ",
The weight of each label is respectively 50%, 20% and 30%, then can determine whether out that the groups of users is badly in need of spending money recently, and according to
Working in Beijing and often going abroad can analyze out the user with certain loan repayment capacity, and the groups of users is to interest at this time
It is high, amount is big and the preference of product that can quickly appropriate funds should be higher.
It is corresponding each to calculate different user reference group according to the reference feature tag of different user reference group by S105
The air control rule match degree of a financial product；
Wherein, the air control rule match degree can be understood as the risk journey that each groups of users buys different financial products
Degree.For example, when the reference feature tag of some groups of users is " top-tier customer ", it can be determined that in the groups of users
User be low-risk client, then the client of low-risk grading can buy the at the same level or higher product of risk class, can also
With the low product of investment risk grade；Similarly, the client of high risk grading may only buy the at the same level or lower production of risk class
Product, and the behavior for investing high-risk grade product will be strictly limited.
In the embodiment of the present invention, as shown in figure 3, the reference feature tag meter according to different user reference group
The method for calculating the air control rule match degree that different user group corresponds to each financial product specifically includes step:
S301 is based on the reference feature tag, different credit angle value is assigned to different user reference group.Such as reference
Feature tag is " top-tier customer ", then credit angle value can assign 80%；Reference feature tag is " medium client ", then credit angle value
" 60% " can be assigned；Reference feature tag is " first-class client ", then credit angle value can assign 90%.
The credit angle value is ranked up by S302；Wherein, the credit angle value is arranged by sequence from big to small
Sequence.
S303, according to the risk class for determining different user reference group that sorts.Wherein, credit angle value is higher, the user
The risk class of group is lower；It is on the contrary then be high risk user.
S106, in conjunction with the air control rule match degree of user behavior group preference and user's reference group to not
Same user carries out financial product recommendation；
It is comprehensive after obtaining the air control rule match degree of preference and different user reference group of different user behavior group
Conjunction analyzes which user is suitble to which kind of financial product, so that corresponding financial product is recommended corresponding user.It lifts
Example explanation, when same user is high to the preference of financial product, while the user is high risk client, can only match risk class
Lower product, therefore system can be to the lower financial product of user's preferential recommendation risk class.
The mode of the recommendation is unlimited, can press recommended products using message push (push), or in browsing pages
User is presented in priority level.
S107, the existing behavioral data based on the user update behavioural characteristic label and the reference spy of the user respectively
Levy label, while redistributing user's owning user group, and calculate update after groups of users preference and air control rule
Then matching degree, Automatic Optimal recommended models.
Wherein, the existing behavioral data of the user include user investment behavior data and borrow refund behavioral data.It is right
In loan product, different user can be tracked in platform where the financial product and borrow refund behavioral data, and borrowed according to this
Refund behavioral data re-execute the steps S102-S107.For finance product, user can be tracked and actually managed money matters now ability,
Recommended it different message again.
Specifically, the real-time tracing user's borrows refund behavioral data, and constantly after user carries out purchase financial product
The behavioural characteristic label and reference feature tag for going to update the user, based on new behavioural characteristic label and reference feature tag weight
The user is newly divided into corresponding user behavior group and user's reference group.For example, being " top-tier customer " before the user
Feature tag, but do not refund on time after practical loaning bill, can update its reference feature tag is " medium client ", and subsequent
Only recommend the lower financial product of risk class for the client.
Financial product intelligent recommendation method provided in an embodiment of the present invention, by generating behavioural characteristic mark according to user information
Label and reference feature tag, and be respectively grouped user according to the behavioural characteristic label and reference feature tag, respectively
Calculate the preference and air control rule match degree of different groups；In conjunction with the preference of different groups and the air control rule of different groups
Matching degree carries out Products Show；Subsequent real-time tracing and update, and Automatic Optimal recommended models.This method is used for different target
Family is marketed different financial products, and dynamic real-time update, the accuracy of recommendation results is improved, so as to more accurately
Recommend the personalized product for being more bonded user itself for user；In addition it also can be improved the marketing efficiency of product.
Meanwhile the embodiment of the present invention provides a kind of intelligent recommendation system of financial product, as shown in Figure 4, comprising:
Acquisition module 41, for acquiring the behavioral data and collage-credit data of user.In the embodiment of the present invention, the behavior
Data include but is not limited to location information and/or application installation preference information of user etc..The location information of the user can base
The location information of mobile terminal user is obtained in location-based service (LBS).The application installation preference information includes in certain time
The installation number of application program and unloading quantity, the installation number of each application program and unloading number etc. in electronic equipment.
The behavioral data of the user can also include click browsing data, browsing duration data, browsing range data of user etc..
The collage-credit data includes personal reference report, public accumulation fund data and social security card data etc..The collage-credit data can
It is acquired by Third Party Authentication.The third party can have for Central Bank's reference center, public accumalation fund for housing construction center, social security office etc. to be recognized
Demonstrate,prove the mechanism of qualification.
Generation module 42 generates behavioural characteristic for being based on the behavioral data and collage-credit data for the user respectively
Label and reference feature tag.It wherein, can be to the behavior when generating the behavioural characteristic label or reference feature tag
Data or collage-credit data carry out word segmentation processing to extract Feature Words.Word segmentation processing is carried out to the behavioral data or collage-credit data
It can be carried out using participle technique well known in the art, the Feature Words extracted are exactly the word obtained after word segmentation processing
Language, or referred to as keyword.
In the embodiment of the present invention, for the behavioural characteristic label, each user has at least one behavioural characteristic mark
Label, behavior feature tag is the keyword that can summarize the main contents of behavioral data of the user.For the sign
Believe feature tag, which is the keyword that can summarize the credit rating of the user.It illustrates, it is assumed that
It is higher that individual subscriber reference report is displayed without overdue record, five one gold medals of danger pay number, then generates " top-tier customer " etc.
Feature tag.
Division module 43, for user to be divided into difference according to the behavioural characteristic label and reference feature tag respectively
User behavior group and different user's reference groups.In the embodiment of the present invention, on the one hand according to the behavioural characteristic label
Corresponding user is divided into different groups, i.e. user behavior group；On the other hand, it will be corresponded to according to the reference feature tag
User be divided into different groups, i.e. user's reference group.Preferably, can by according to the keyword of the feature tag whether
It is similar or close, similar or similar user is classified as one kind.
Preference computing module 44, for calculating different use according to the behavioural characteristic label of different user behavior group
Family behavior group corresponds to the preference of each financial product.Wherein, the financial product can be understood as including different financings
Product or different loan products.The preference of the user behavior group can be understood as the use in the different user group
Degree of concern of the family to financial product.Higher to certain a finance product degree of concern, the user represented in the groups of users throws
It is very high to provide purchase intention；Degree of concern is not high, represents user in the groups of users not no intention of Investment & Financing recently.
In the embodiment of the present invention, the preference computing module 44 is also specifically included:
Weight determination module 441, the weight of the different behavioural characteristic labels for determining same user behavior group.Its
In, the assignment of weight can be carried out according to the probability that the Feature Words of the different behavioural characteristic labels occur.
Weight sequencing module 442, for sorting from large to small different behavioural characteristic label weights；
Weight selecting module 443, for using the weight limit of behavioural characteristic label as the user behavior group to finance
The preference of product.
For example, if the feature tag of a groups of users includes " loan ", " being located at Beijing " and " travel abroad ",
The weight of each label is respectively 50%, 20% and 30%, then can determine whether out that the groups of users is badly in need of spending money recently, and according to
Working in Beijing and often going abroad can analyze out the user with certain loan repayment capacity, and the groups of users is to interest at this time
It is high, amount is big and the preference of product that can quickly appropriate funds should be higher.
Air control matching degree computing module 45, for being calculated not according to the reference feature tag of different user reference group
The air control rule match degree of each financial product is corresponded to user's reference group.Wherein, the air control rule match degree can be managed
Solution is the degree of risk that each groups of users buys different financial products.For example, the reference when some groups of users is special
When sign label is " top-tier customer ", it can be determined that the user in the groups of users is low-risk client, then the low-risk is graded
Client can buy the at the same level or higher product of risk class, product that can also be low with investment risk grade；Similarly, high risk is commented
The client of grade may only buy the at the same level or lower product of risk class, and the behavior for investing high-risk grade product will be tight
Lattice limitation.
In the embodiment of the present invention, the air control matching degree computing module 45 is also specifically included:
Credit assignment module 451 assigns different for being based on the reference feature tag to different user reference group
Credit angle value.If reference feature tag is " top-tier customer ", then credit angle value can assign 80%；Reference feature tag is " medium
Client ", then credit angle value can assign " 60% "；Reference feature tag is " first-class client ", then credit angle value can assign 90%.
Credit sorting module 452, for the credit angle value to be ranked up；It wherein, will be described by sequence from big to small
Credit angle value is ranked up.
Risk determining module 453, for according to the risk class for determining different user reference group that sorts.Wherein, credit
Angle value is higher, and the risk class of the groups of users is lower；It is on the contrary then be high risk user.
Recommending module 46, for the air control rule in conjunction with user behavior group preference and user's reference group
Matching degree carries out financial product recommendation to different users.In the preference and different user sign for obtaining different user behavior group
After the air control rule match degree for believing group, comprehensive analysis goes out which user is suitble to which kind of financial product, thus by corresponding
Financial product recommends corresponding user.For example, when preference of the same user to financial product is high, while the user is
High risk client, can only match the lower product of risk class, thus system can to user's preferential recommendation risk class compared with
Low financial product.
The mode of the recommendation is unlimited, can press recommended products using message push (push), or in browsing pages
User is presented in priority level.
Update module 47 updates the behavioural characteristic mark of the user for the existing behavioral data based on the user respectively
Label and reference feature tag, while user's owning user group is redistributed, and calculate the preference of groups of users after update
Degree and air control rule match degree, Automatic Optimal recommended models.Wherein, the existing behavioral data of the user includes the investment of user
Behavioral data and borrow refund behavioral data.For loan product, different user can be tracked in platform where the financial product
Borrow refund behavioral data, and according to this borrow refund behavioral data re-execute the steps S102-S107.It, can for finance product
It is actually managed money matters now ability with tracking user, is recommended it different message again.
Specifically, the real-time tracing user's borrows refund behavioral data, and constantly after user carries out purchase financial product
The behavioural characteristic label and reference feature tag for going to update the user, based on new behavioural characteristic label and reference feature tag weight
The user is newly divided into corresponding user behavior group and user's reference group.For example, being " top-tier customer " before the user
Feature tag, but do not refund on time after practical loaning bill, can update its reference feature tag is " medium client ", and subsequent
Only recommend the lower financial product of risk class for the client.
Financial product intelligent recommendation system provided in an embodiment of the present invention, by generating behavioural characteristic mark according to user information
Label and reference feature tag, and be respectively grouped user according to the behavioural characteristic label and reference feature tag, respectively
Calculate the preference and air control rule match degree of different groups；In conjunction with the preference of different groups and the air control rule of different groups
Matching degree carries out Products Show；Subsequent real-time tracing and update, and Automatic Optimal recommended models.This method is used for different target
Family is marketed different financial products, and dynamic real-time update, the accuracy of recommendation results is improved, so as to more accurately
Recommend the personalized product for being more bonded user itself for user；In addition it also can be improved the marketing efficiency of product.
Although the preferred embodiment of the application has been described, it is created once a person skilled in the art knows basic
Property concept, then additional changes and modifications may be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as
It selects embodiment and falls into all change and modification of the application range.
Obviously, those skilled in the art can carry out various modification and variations without departing from the essence of the application to the application
Mind and range.In this way, if these modifications and variations of the application belong to the range of the claim of this application and its equivalent technologies
Within, then the application is also intended to include these modifications and variations.
Claims (10)
1. a kind of intelligent recommendation method of financial product characterized by comprising
Acquire the behavioral data and collage-credit data of user；
Based on the behavioral data and collage-credit data, behavioural characteristic label and reference feature tag are generated respectively for the user；
User is divided by different user behavior group and not according to the behavioural characteristic label and reference feature tag respectively
Same user's reference group；
Different user behavior group, which is calculated, according to the behavioural characteristic label of different user behavior group corresponds to each finance production
The preference of product；
Different user reference group, which is calculated, according to the reference feature tag of different user reference group corresponds to each finance production
The air control rule match degree of product；
In conjunction with the air control rule match degree of user behavior group preference and user's reference group to different users
Carry out financial product recommendation；
Existing behavioral data based on the user updates the behavioural characteristic label and reference feature tag of the user respectively, together
When redistribute user's owning user group, and calculate update after groups of users preference and air control rule match degree,
Automatic Optimal recommended models.
2. intelligent recommendation method as described in claim 1, which is characterized in that the behavioral data includes the location information of user
And/or application installation preference information.
3. intelligent recommendation method as described in claim 1, which is characterized in that the collage-credit data include personal reference report,
Public accumulation fund data and social security card data.
4. intelligent recommendation method as described in claim 1, which is characterized in that generating the behavioural characteristic label or reference spy
When levying label, word segmentation processing is carried out to extract Feature Words to the behavioral data or collage-credit data.
5. intelligent recommendation method as described in claim 1, which is characterized in that according to the behavioural characteristic of different user behavior group
The method that label calculates the preference that different user behavior group corresponds to each financial product specifically includes:
Determine the weight of the different behavioural characteristic labels of same user behavior group；
Different behavioural characteristic label weights is sorted from large to small；
Using the weight limit of behavioural characteristic label as the user behavior group to the preference of financial product.
6. intelligent recommendation method as described in claim 1, which is characterized in that described according to different user reference group
The method that reference feature tag calculates the air control rule match degree that different user group corresponds to each financial product specifically includes:
Based on the reference feature tag, different credit angle value is assigned to different user reference group；
The credit angle value is ranked up；
According to the risk class for determining different user reference group that sorts.
7. intelligent recommendation method as described in claim 1, which is characterized in that the existing behavioral data of the user includes user
Investment behavior data and borrow refund behavioral data.
8. a kind of intelligent recommendation system of financial product characterized by comprising
Acquisition module, for acquiring the behavioral data and collage-credit data of user；
Generation module, for be based on the behavioral data and collage-credit data, for the user generate respectively behavioural characteristic label and
Reference feature tag；
Division module, for user to be divided into different users according to the behavioural characteristic label and reference feature tag respectively
Behavior group and different user's reference groups；
Preference computing module, for calculating different user behavior according to the behavioural characteristic label of different user behavior group
Group corresponds to the preference of each financial product；
Air control matching degree computing module, for calculating different user according to the reference feature tag of different user reference group
Reference group corresponds to the air control rule match degree of each financial product；
Recommending module, for the air control rule match degree in conjunction with user behavior group preference and user's reference group
Financial product recommendation is carried out to different users；
Update module updates the behavioural characteristic label and sign of the user for the existing behavioral data based on the user respectively
Believe feature tag, while redistributing user's owning user group, and calculates the preference and wind of groups of users after update
Regulatory control then matching degree, Automatic Optimal recommended models.
9. the intelligent recommendation system as shown in claim 8, which is characterized in that the behavioral data includes the positioning of user
Information and/or application installation preference information.
10. the intelligent recommendation system as shown in claim 8, which is characterized in that the existing behavioral data of the user includes using
The investment behavior data at family and borrow refund behavioral data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810939473.6A CN109325845A (en) | 2018-08-15 | 2018-08-15 | A kind of financial product intelligent recommendation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810939473.6A CN109325845A (en) | 2018-08-15 | 2018-08-15 | A kind of financial product intelligent recommendation method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109325845A true CN109325845A (en) | 2019-02-12 |
Family
ID=65263270
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810939473.6A Pending CN109325845A (en) | 2018-08-15 | 2018-08-15 | A kind of financial product intelligent recommendation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109325845A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110287408A (en) * | 2019-05-23 | 2019-09-27 | 上海拍拍贷金融信息服务有限公司 | A method and system for customizing and pushing financial products |
| WO2020164269A1 (en) * | 2019-02-15 | 2020-08-20 | 平安科技（深圳）有限公司 | User-group-based message pushing method and apparatus, and computer device |
| CN111815440A (en) * | 2020-09-14 | 2020-10-23 | 杭州连银科技有限公司 | A loan supermarket marketing data processing system |
| CN112330404A (en) * | 2020-11-10 | 2021-02-05 | 广发证券股份有限公司 | Data processing method and device, server and storage medium |
| CN112508685A (en) * | 2020-12-08 | 2021-03-16 | 苏州方正璞华信息技术有限公司 | Enterprise credit investigation information-based labeling method and system |
| CN113469822A (en) * | 2021-06-03 | 2021-10-01 | 北京优全智汇信息技术有限公司 | Method and system for risk prevention, control and early warning in insurance industry |
| CN113487380A (en) * | 2021-06-25 | 2021-10-08 | 天元大数据信用管理有限公司 | Financial product recommendation method, device, equipment and medium |
| CN114757782A (en) * | 2022-04-21 | 2022-07-15 | 深圳极联信息技术股份有限公司 | Intelligent financial service system based on big data |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104156356A (en) * | 2013-05-13 | 2014-11-19 | 腾讯科技（深圳）有限公司 | Personalized navigation page generating method and device |
| CN107194515A (en) * | 2017-05-27 | 2017-09-22 | 冯小平 | Determine user's current behavior and the method and apparatus for predicting user view |
| CN107784066A (en) * | 2017-08-25 | 2018-03-09 | 上海壹账通金融科技有限公司 | Information recommendation method, device, server and storage medium |
| CN107993146A (en) * | 2018-01-25 | 2018-05-04 | 深圳市前海吉顺信科技发展有限公司 | The air control method and system of financial big data |
-
2018
- 2018-08-15 CN CN201810939473.6A patent/CN109325845A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104156356A (en) * | 2013-05-13 | 2014-11-19 | 腾讯科技（深圳）有限公司 | Personalized navigation page generating method and device |
| CN107194515A (en) * | 2017-05-27 | 2017-09-22 | 冯小平 | Determine user's current behavior and the method and apparatus for predicting user view |
| CN107784066A (en) * | 2017-08-25 | 2018-03-09 | 上海壹账通金融科技有限公司 | Information recommendation method, device, server and storage medium |
| CN107993146A (en) * | 2018-01-25 | 2018-05-04 | 深圳市前海吉顺信科技发展有限公司 | The air control method and system of financial big data |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020164269A1 (en) * | 2019-02-15 | 2020-08-20 | 平安科技（深圳）有限公司 | User-group-based message pushing method and apparatus, and computer device |
| CN110287408A (en) * | 2019-05-23 | 2019-09-27 | 上海拍拍贷金融信息服务有限公司 | A method and system for customizing and pushing financial products |
| CN111815440A (en) * | 2020-09-14 | 2020-10-23 | 杭州连银科技有限公司 | A loan supermarket marketing data processing system |
| CN111815440B (en) * | 2020-09-14 | 2020-12-08 | 杭州连银科技有限公司 | A loan supermarket marketing data processing system |
| CN112330404A (en) * | 2020-11-10 | 2021-02-05 | 广发证券股份有限公司 | Data processing method and device, server and storage medium |
| CN112508685A (en) * | 2020-12-08 | 2021-03-16 | 苏州方正璞华信息技术有限公司 | Enterprise credit investigation information-based labeling method and system |
| CN113469822A (en) * | 2021-06-03 | 2021-10-01 | 北京优全智汇信息技术有限公司 | Method and system for risk prevention, control and early warning in insurance industry |
| CN113487380A (en) * | 2021-06-25 | 2021-10-08 | 天元大数据信用管理有限公司 | Financial product recommendation method, device, equipment and medium |
| CN114757782A (en) * | 2022-04-21 | 2022-07-15 | 深圳极联信息技术股份有限公司 | Intelligent financial service system based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109325845A (en) | A kind of financial product intelligent recommendation method and system | |
| CN106447434A (en) | Personal credit ecological platform | |
| CN106503873A (en) | A kind of prediction user follows treaty method, device and the computing device of probability | |
| CN109035003A (en) | Anti- fraud model modelling approach and anti-fraud monitoring method based on machine learning | |
| CN105490823B (en) | data processing method and device | |
| CN114119137A (en) | Risk control method and device | |
| CN102082995A (en) | Method, system and device for providing positional information | |
| CN110097450A (en) | Vehicle borrows methods of risk assessment, device, equipment and storage medium | |
| CN110276677A (en) | Refund prediction technique, device, equipment and storage medium based on big data platform | |
| CN117993910A (en) | Verification method and related device for abnormal transaction response strategy | |
| US10785594B2 (en) | Network based resource management and allocation | |
| US20220335485A1 (en) | Partner fee recommendation service | |
| KR102630611B1 (en) | System for providing map based franchise foundation simulation service using market analysis | |
| CN111062770B (en) | Merchant identification method, device and computer readable medium | |
| CN108022150A (en) | recommendation method and system based on O2O data | |
| CN110858326A (en) | Method, device, equipment and medium for model training and acquiring additional characteristic data | |
| CN109727125A (en) | Loan amount prediction method, device, server and storage medium | |
| US20250078157A1 (en) | Asset-backed tokenization system for progressive asset ownership | |
| CN111476657B (en) | Information pushing method, device and system | |
| CN109191185A (en) | A kind of visitor's heap sort method and system | |
| CN113327161A (en) | Intelligent wind control decision system for credit business | |
| CN111178535B (en) | Method and apparatus for implementing automatic machine learning | |
| CN111429251A (en) | Method and device for processing data under multiple modes | |
| CN116542696A (en) | User group classification method, device, equipment and storage medium | |
| CN112767002A (en) | Block chain-based travel operator environment and travel area evaluation method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190212 |