CN109300050A - Insurance method for pushing, device and storage medium based on user's portrait - Google Patents

Insurance method for pushing, device and storage medium based on user's portrait Download PDF

Info

Publication number
CN109300050A
CN109300050A CN201811013020.7A CN201811013020A CN109300050A CN 109300050 A CN109300050 A CN 109300050A CN 201811013020 A CN201811013020 A CN 201811013020A CN 109300050 A CN109300050 A CN 109300050A
Authority
CN
China
Prior art keywords
insurance
user
transaction
portrait
historical trading
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
Application number
CN201811013020.7A
Other languages
Chinese (zh)
Inventor
顾宝宝
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ping An Technology Shenzhen Co Ltd
Original Assignee
Ping An Technology Shenzhen Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Ping An Technology Shenzhen Co Ltd filed Critical Ping An Technology Shenzhen Co Ltd
Priority to CN201811013020.7A priority Critical patent/CN109300050A/en
Publication of CN109300050A publication Critical patent/CN109300050A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08Insurance, e.g. risk analysis or pensions

Abstract

The embodiment of the invention provides a kind of insurance method for pushing, device and storage mediums based on user's portrait.The historical trading data that the embodiment of the present invention passes through acquisition user, then, utilize machine learning algorithm and feature extraction algorithm, the historical trading data is handled, the insurance transaction portrait of the user is obtained, the insurance transaction portrait is inclined to degree for the transaction of each insurance to characterize the user, thus, it is traded and is drawn a portrait according to the insurance, pushed and insure to the user.Therefore, the mode that technical solution provided in an embodiment of the present invention is able to solve artificial push insurance in the prior art there are problems that recommending accuracy poor and heavy workload.

Description

Insurance method for pushing, device and storage medium based on user's portrait
[technical field]
The present invention relates to field of computer technology more particularly to a kind of insurance method for pushing based on user's portrait, device And storage medium.
[background technique]
Currently, when user need buy insurance when, generally by insurance sales personnel according to the demand of user and work Experience, judged in a manner of subjective judgement user be more prone to transaction insurance type, and by these insurance types to user into Row explanation is introduced.Based on this, insurance in the prior art, which is introduced, carrys out artificial judgment completion generally by insurance sales personnel.
But the subjective judgement of insurance sales personnel is mainly subject to by the safe manner of artificial judgment user demand, This biggish experience and subjective judgement dependent on insurance sales personnel, while heavy workload, due to the guarantor to user demand The accuracy of judgement degree of danger is poor, this just needs to repeat to judge to recommend, and further increases workload.
[summary of the invention]
In view of this, the embodiment of the invention provides a kind of insurance method for pushing, device and storages based on user's portrait Medium recommends accuracy poor and heavy workload to solve the problems, such as that the mode for manually pushing insurance in the prior art exists.
In a first aspect, the embodiment of the invention provides a kind of insurance method for pushing based on user's portrait, comprising:
Obtain the historical trading data of user;
Using machine learning algorithm and feature extraction algorithm, the historical trading data is handled, obtains the use The insurance at family, which is traded, draws a portrait, and the insurance transaction portrait is to characterize the user for the transaction tendency degree of each insurance;
It is traded and is drawn a portrait according to the insurance, pushed and insure to the user.
The aspect and any possible implementation manners as described above, it is further provided a kind of implementation utilizes engineering Algorithm and feature extraction algorithm are practised, the historical trading data is handled, obtains the insurance transaction portrait of the user, packet It includes:
The historical trading data is handled using the feature extraction algorithm, the history for obtaining the historical trading data is handed over Easy feature and transaction tendency feature;
Neural network machine is utilized using transaction tendency feature as output using the historical trading feature as input System carries out deep learning, obtains insurance transaction probability prediction model;
Each insurance is predicted by the insurance transaction probability prediction model, obtains the user for each insurance Transaction tendency degree, is drawn a portrait using trading as the insurance.
The aspect and any possible implementation manners as described above, it is further provided a kind of implementation, it is described will be described Historical trading feature is as input, using transaction tendency feature as output, carries out deep learning using neural network mechanism, Obtain insurance transaction probability prediction model, comprising:
It is calculated using transaction tendency feature as output using random forest using the historical trading feature as input At least one of method and decision tree machine algorithm carry out deep learning, obtain the insurance transaction probability prediction model.
The aspect and any possible implementation manners as described above, it is further provided a kind of implementation, it is described according to institute Insurance transaction portrait is stated, pushes and insures to the user, comprising:
It is traded and is drawn a portrait according to the insurance, determine that target is insured;
Target insurance is pushed to the user.
The aspect and any possible implementation manners as described above, it is further provided a kind of implementation, it is described according to institute Insurance transaction portrait is stated, determines that target is insured, comprising:
It is directed to the sequence of the transaction tendency degree of each insurance from high to low according to the user, each insurance is ranked up;
The insurance of specifying number from front to back is obtained in the sequence, it is described specified to insure as the target Number is the integer greater than 0.
The aspect and any possible implementation manners as described above, it is further provided a kind of implementation, it is described will be described Target insurance is pushed to the user, comprising:
Target insurance list is pushed to the user;Alternatively, being sequentially output each target insurance.
A technical solution in above-mentioned technical proposal has the following beneficial effects:
It in the embodiment of the present invention, is analyzed by the historical trading data to user, and based on to these historical tradings Data carry out the processing of feature extraction and machine learning algorithm, can obtain meeting user for the guarantor of the transaction tendency of each insurance Danger transaction portrait, is based on this, the insurance for more meeting customer transaction and being inclined to can be pushed to user, this is going through based on each user What history transaction data was made judges automatically, and compared to the mode of insurance personnel's artificial judgment, can improve and push away to a certain extent Accuracy is recommended, and this workload that automatically processes mode also and can reduce insurance recommendation personnel, improves its working efficiency.
Second aspect, the embodiment of the invention provides a kind of insurance driving means based on user's portrait, comprising:
Acquiring unit, for obtaining the historical trading data of user;
Processing unit, for utilizing machine learning algorithm and feature extraction algorithm, at the historical trading data Reason obtains the insurance transaction portrait of the user, the friendship that the insurance transaction portrait is directed to each insurance to characterize the user Easily tendency degree;
Push unit is drawn a portrait for being traded according to the insurance, is pushed and is insured to the user.
The aspect and any possible implementation manners as described above, it is further provided a kind of implementation, the processing are single Member is specifically used for:
The historical trading data is handled using the feature extraction algorithm, the history for obtaining the historical trading data is handed over Easy feature and transaction tendency feature;
Neural network machine is utilized using transaction tendency feature as output using the historical trading feature as input System carries out deep learning, obtains insurance transaction probability prediction model;
Each insurance is predicted by the insurance transaction probability prediction model, obtains the user for each insurance Transaction tendency degree, is drawn a portrait using trading as the insurance.
The aspect and any possible implementation manners as described above, it is further provided a kind of implementation, the processing are single Member, also particularly useful for:
It is calculated using transaction tendency feature as output using random forest using the historical trading feature as input At least one of method and decision tree machine algorithm carry out deep learning, obtain the insurance transaction probability prediction model.
The aspect and any possible implementation manners as described above, it is further provided a kind of implementation, the push are single Member is used for:
It is traded and is drawn a portrait according to the insurance, determine that target is insured;
Target insurance is pushed to the user.
The aspect and any possible implementation manners as described above, it is further provided a kind of implementation, the push are single Member is specifically used for:
It is directed to the sequence of the transaction tendency degree of each insurance from high to low according to the user, each insurance is ranked up;
The insurance of specifying number from front to back is obtained in the sequence, it is described specified to insure as the target Number is the integer greater than 0.
The aspect and any possible implementation manners as described above, it is further provided a kind of implementation, the push are single Member is specifically used for:
Target insurance list is pushed to the user;Alternatively, being sequentially output each target insurance.
A technical solution in above-mentioned technical proposal has the following beneficial effects:
It in the embodiment of the present invention, is analyzed by the historical trading data to user, and based on to these historical tradings Data carry out the processing of feature extraction and machine learning algorithm, can obtain meeting user for the guarantor of the transaction tendency of each insurance Danger transaction portrait, is based on this, the insurance for more meeting customer transaction and being inclined to can be pushed to user, this is going through based on each user What history transaction data was made judges automatically, and compared to the mode of insurance personnel's artificial judgment, can improve and push away to a certain extent Accuracy is recommended, and this workload that automatically processes mode also and can reduce insurance recommendation personnel, improves its working efficiency.
The third aspect, the embodiment of the invention provides a kind of insurance driving means based on user's portrait, which is characterized in that Include: memory, processor and storage in the memory and the computer program that can run on the processor, It is characterized in that, the processor is realized when executing the computer program such as the step of any one of first aspect the method.
Fourth aspect, the embodiment of the invention provides a kind of storage mediums characterized by comprising computer is executable Instruction is drawn as first aspect is described in any item based on user when the computer executable instructions are run to execute The insurance method for pushing of picture.
A technical solution in above-mentioned technical proposal has the following beneficial effects:
It in the embodiment of the present invention, is analyzed by the historical trading data to user, and based on to these historical tradings Data carry out the processing of feature extraction and machine learning algorithm, can obtain meeting user for the guarantor of the transaction tendency of each insurance Danger transaction portrait, is based on this, the insurance for more meeting customer transaction and being inclined to can be pushed to user, this is going through based on each user What history transaction data was made judges automatically, and compared to the mode of insurance personnel's artificial judgment, can improve and push away to a certain extent Accuracy is recommended, and this workload that automatically processes mode also and can reduce insurance recommendation personnel, improves its working efficiency.
[Detailed description of the invention]
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached Figure is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this field For those of ordinary skill, without any creative labor, it can also be obtained according to these attached drawings other attached Figure.
Fig. 1 is that the process of the embodiment one of the insurance method for pushing based on user's portrait provided by the embodiment of the present invention is shown It is intended to;
Fig. 2 is that the process of the embodiment two of the insurance method for pushing based on user's portrait provided by the embodiment of the present invention is shown It is intended to;
Fig. 3 is that the process of the embodiment three of the insurance method for pushing based on user's portrait provided by the embodiment of the present invention is shown It is intended to;
Fig. 4 is the functional block diagram of the insurance driving means provided by the embodiment of the present invention based on user's portrait;
Fig. 5 is the entity apparatus schematic diagram of the insurance driving means provided by the embodiment of the present invention based on user's portrait.
[specific embodiment]
For a better understanding of the technical solution of the present invention, being retouched in detail to the embodiment of the present invention with reference to the accompanying drawing It states.
It will be appreciated that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Base Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts it is all its Its embodiment, shall fall within the protection scope of the present invention.
The term used in embodiments of the present invention is only to be not intended to be limiting merely for for the purpose of describing particular embodiments The present invention.In the embodiment of the present invention and the "an" of singular used in the attached claims, " described " and "the" It is also intended to including most forms, unless the context clearly indicates other meaning.
It should be appreciated that term "and/or" used herein is only a kind of incidence relation for describing affiliated partner, indicate There may be three kinds of relationships, for example, A and/or B, can indicate: individualism A, exist simultaneously A and B, individualism B these three Situation.In addition, character "/" herein, typicallys represent the relationship that forward-backward correlation object is a kind of "or".
Depending on context, word as used in this " if " can be construed to " ... when " or " when ... When " or " in response to determination " or " in response to detection ".Similarly, depend on context, phrase " if it is determined that " or " if detection (condition or event of statement) " can be construed to " when determining " or " in response to determination " or " when the detection (condition of statement Or event) when " or " in response to detection (condition or event of statement) ".
For in the presence of the prior art by way of artificial judgment user's demand for insurance there are accuracy of judgement degree compared with The problem of difference and heavy workload, the embodiment of the invention provides following resolving ideas: the historical trading data of user is obtained, to this A little transaction data are handled, and are inclined to degree to obtain user for the transaction of each insurance, form the insurance transaction portrait of user, Since insurance transaction portrait is what the historical trading data based on user obtained, the habit of transaction and transaction for being more in line with user are inclined To, it is based on this, when pushing insurance for user, push accuracy rate can be improved to a certain extent, it also can be to a certain extent Transaction rate is improved, and liberates manpower.
Under the guidance of the thinking, this programme embodiment provides following feasible embodiment.
Embodiment one
The embodiment of the present invention provides a kind of insurance method for pushing based on user's portrait.It can with specific reference to Fig. 1, this method To include the following steps:
S102 obtains the historical trading data of user.
In the embodiment of the present invention, historical trading data includes history insurance transaction data, wherein related history insurance Transaction data is transaction data relevant to insurance business.Specifically, can include but is not limited to: what user bought before Vehicle insurance transaction data, the life insurance transaction data bought between user, etc..
These history insurance transaction data can be looked by way of the history declaration form stored in inquiry own database Inquiry obtains.Alternatively, can also be obtained by other approach.For example, user actively provides;For example, such as being passed through by other channels Insurance industry block chain etc. obtains the disclosed insurance Transaction Information of user.
In the scene of practical application, historical trading data can also include be able to reflect customer transaction habit other go through History Transaction Information.In some possible realization scenes, otherwise information be can include but is not limited to: history purchase transaction At least one of data, history loan transaction data and history fixed assets transaction data.Above-mentioned other historical tradings letter The acquisition modes of breath are then not particularly limited.For example, user actively provides;For example, such as passing through trading industry area by other channels Block chain etc. obtains the disclosed historical transactional information of user.
S104 is handled historical trading data using machine learning algorithm and feature extraction algorithm, obtains user's Insurance transaction portrait, insurance transaction portrait are inclined to degree for the transaction of each insurance to characterize user.
It should be noted that insurance transaction portrait can be there are many form of expression involved in the embodiment of the present invention.For example, Friendship of the user to all kinds of insurances (such as: personal class insurance and the different classes of insurance of vehicle insurance) can be characterized with the preset form of expression Easily tendency degree;Alternatively, in another example, user can also be characterized with the preset form of expression to be insured (such as: protecting for personal class for each Each insurance of dangerous subordinate) transaction be inclined to degree.Wherein, preset performance shape of the embodiment of the present invention for tendency degree of trading Formula is not particularly limited.For example, can with representative fraction, such as percentage, mode embody, alternatively, can also just be inclined to and negative tendency Mode it is embodied, alternatively, can be with.In actual use, it is set as needed.
S106 is pushed to user and is insured according to insurance transaction portrait.
Based on the historical transactional information got in S102, the embodiment of the present invention provides the realization of S104 step performed below Mode, referring to FIG. 2, S104 specifically comprises the following steps:
S1042 handles the historical trading data using feature extraction algorithm, obtains the history of the historical trading data Transaction feature and transaction tendency feature.
Wherein, feature extraction algorithm involved in S1042 step can include but is not limited to: N-Gram algorithm.
For example, can be carried out to it if historical trading data is " purchase accident insurance X member " based on N-Gram algorithm Feature extraction obtains corresponding Unigram feature, Bigram feature, Trigram feature etc.;Wherein, each Unigram feature Comprising a word, each Bigram feature includes two words, and each Trigram feature includes three words.Based on this, above-mentioned history The historical trading feature that transaction data can extract after N-Gram algorithm process can include but is not limited to: purchase, buy, anticipating, Outside, danger, X, member, purchase, accident, X member and accident insurance, etc.;The form of expression of obtained transaction tendency feature can be general for transaction Rate, alternatively, be positive tendency and countertendency can also be showed simply, wherein positive trend has transaction to be inclined to, and countertendency indicates No deal tendency.
In concrete implementation scene, the historical trading feature obtained using feature extraction algorithm may include but unlimited In: the purchase amount of money, purchase frequency, insurance kind of purchase etc..
S1044 utilizes neural network mechanism using transaction tendency feature as output using historical trading feature as input Deep learning is carried out, insurance transaction probability prediction model is obtained.
S1046 predicts each insurance by insuring transaction probability prediction model, obtains the friendship that user is directed to each insurance Easily tendency degree, to draw a portrait as insurance transaction.
Wherein, machine learning algorithm involved in S1044 step can include but is not limited to: decision tree machine algorithm with Machine forest machine algorithm two ways is realized.
The first, executes S1044 step by decision tree machine algorithm: using historical trading feature as input, will trade Feature is inclined to as output, carries out deep learning using decision tree machine algorithm, obtains insurance transaction probability prediction model.
Wherein, decision tree (Decision Tree) is a kind of probability analysis method, in machine learning, decision tree conduct One prediction model, for characterizing a kind of mapping relations between object symbolic animal of the birth year and object value.Each node in tree indicates One object, and some possible attribute value that each diverging paths then represent, and each leaf node then correspond to from root node to The value of object represented by the path of leaf node experience.In the embodiment of the present invention, can using the historical trading data of user as The preset decision-tree model of sample data treatment is trained by sample data, and the insurance transaction probability that user can be obtained is pre- Survey model.
Therefore, historical trading feature is handled by decision tree machine algorithm, the insurance transaction of user can be obtained Probabilistic model, the insurance transaction probability model can be realized user for the transaction tendency degree of insurance.In this realization scene In, it is subsequent when implementing S1046, the insurance transaction probability prediction model can be utilized respectively, the transaction of each insurance is inclined to Degree is predicted.
In this implementation, by decision tree machine algorithm obtain user insurance trade portrait, should be readily appreciated that and It realizes, also, lower to the processing requirement of sample data, can realize the analysis to big data in a relatively short period of time, analyze Accuracy is higher.
Second, S1044 is executed using historical trading feature as input by random forests algorithm, feature is inclined in transaction As output, deep learning is carried out using random forests algorithm, obtains insurance transaction probability prediction model.
Wherein, random forest belongs to the bagging bagging algorithm in integrated study (Ensemble Learning), mainly For the combination algorithm of bagging algorithm and decision Tree algorithms.Wherein, the algorithmic procedure of bagging is substantially are as follows: from original sample collection It is middle to randomly select n training sample using Bootstraping method, k wheel is carried out altogether and is extracted, and k training set is obtained.(k instruction Practice between collection independently of each other, element can have repetition), later, for k training set, we train k model (this k model Can be depending on particular problem, such as decision tree, knn etc.), thus, for classification problem: generating classification knot by voting Fruit;For regression problem: by the mean value of k model prediction result as last prediction result.
It, can be in inside for one when building forest since random forests algorithm can handle a large amount of input varible As error after change generate the estimation of not deviation, also, this machine algorithm is for unbalanced grouped data collection, can be with Balance error, it is more accurate that the user that these features obtain random forests algorithm insures transaction portrait;And due to study speed Degree is very fast, and data-handling efficiency is higher.
It, can be with the combination of a kind of above-mentioned implementation or at least two implementations when implementing S1044 To execute S104.
Based on this, when executing S106, then Fig. 3 can be referred to, comprising the following steps:
S1062 determines that target is insured according to insurance transaction portrait.
Wherein, the number of the insurance of target determined by the step is at least one.For example, in some scenes, in order to expand Big probability of transaction can also determine multiple target insurances.
Target insurance is pushed to user by S1064.
Wherein, it executes in S1062 step according to insurance transaction portrait, determines the mode of target insurance, may include following Step:
It is directed to the sequence of the transaction tendency degree of each insurance from high to low according to user, each insurance is ranked up;
The insurance that specifies number obtained in the ranking from front to back is specified number using insuring as target as greater than 0 Integer.
Specifically, specifying number can be set as needed.If specifying number is one, sequence is obtained near preceding one A insurance is insured as target;Alternatively, when specify number as N, and when N is the integer greater than 1, then it is forward to obtain sequence N number of insurance is insured as target.
In this way, acquisition and determination to target insurance can be realized.
Based on this, when executing S1064 step, then target can be insured into list and be pushed to user;Alternatively, being sequentially output Each target insurance.
The technical solution of the embodiment of the present invention has the advantages that
It in the embodiment of the present invention, is analyzed by the historical trading data to user, and based on to these historical tradings Data carry out the processing of feature extraction and machine learning algorithm, can obtain meeting user for the guarantor of the transaction tendency of each insurance Danger transaction portrait, is based on this, the insurance for more meeting customer transaction and being inclined to can be pushed to user, this is going through based on each user What history transaction data was made judges automatically, and compared to the mode of insurance personnel's artificial judgment, can improve and push away to a certain extent Accuracy is recommended, and this workload that automatically processes mode also and can reduce insurance recommendation personnel, improves its working efficiency.
Embodiment two
The insurance method for pushing based on user's portrait provided by one, the embodiment of the present invention are further based on the above embodiment It provides and realizes the Installation practice of each step and method in above method embodiment.
On the one hand, the embodiment of the invention provides a kind of insurance driving means based on user's portrait.Specifically, can join Fig. 4 is examined, should include: based on the insurance driving means 400 that user draws a portrait
Acquiring unit 41, for obtaining the historical trading data of user;
Processing unit 42, for handling historical trading data using machine learning algorithm and feature extraction algorithm, The insurance for obtaining user, which is traded, draws a portrait, and insurance transaction portrait is to characterize user for the transaction tendency degree of each insurance;
Push unit 43, for pushing and insuring to user according to insurance transaction portrait.
Wherein, processing unit 42 are specifically used for:
The historical trading data is handled using the feature extraction algorithm, the history for obtaining the historical trading data is handed over Easy feature and transaction tendency feature;
Neural network machine is utilized using transaction tendency feature as output using the historical trading feature as input System carries out deep learning, obtains insurance transaction probability prediction model;
Each insurance is predicted by the insurance transaction probability prediction model, obtains the user for each insurance Transaction tendency degree, is drawn a portrait using trading as the insurance.
During a concrete implementation, the processing unit 42, also particularly useful for:
It is calculated using transaction tendency feature as output using random forest using the historical trading feature as input At least one of method and decision tree machine algorithm carry out deep learning, obtain the insurance transaction probability prediction model.
In the embodiment of the present invention, the push unit 43 is used for:
It is traded and is drawn a portrait according to the insurance, determine that target is insured;
Target insurance is pushed to the user.
On the one hand, it according to insurance transaction portrait, determines in the step for target is insured, push unit 43 can be used specifically In:
It is directed to the sequence of the transaction tendency degree of each insurance from high to low according to user, each insurance is ranked up;
The insurance that specifies number obtained in the ranking from front to back is specified number using insuring as target as greater than 0 Integer.
On the other hand, by target insurance be pushed to user the step of in, push unit 43 can include but is not limited to Lower processing mode:
Target insurance list is pushed to user;Alternatively,
It is sequentially output each target insurance.
Historical trading data involved in the embodiment of the present invention, further includes: history purchase transaction data, history loan transaction At least one of data and history fixed assets transaction data.
In addition, the embodiment of the invention also provides a kind of answering devices based on deep learning, referring to FIG. 5, this is based on The insurance driving means 500 of user's portrait includes: memory 51, processor 52 and is stored in memory 51 and can handle The computer program run on device 52, processor 52 are realized as described in one any one of embodiment when executing computer program based on use The step of insurance method for pushing of family portrait.
On the other hand, the embodiment of the invention provides a kind of storage mediums, comprising: computer executable instructions, when described When computer executable instructions are run, to execute the insurance push drawn a portrait as embodiment one is described in any item based on user Method.
Method shown in embodiment one is able to carry out as each unit in this present embodiment, what the present embodiment was not described in detail Part can refer to the related description to embodiment one.
The technical solution of the embodiment of the present invention has the advantages that
It in the embodiment of the present invention, is analyzed by the historical trading data to user, and based on to these historical tradings Data carry out the processing of feature extraction and machine learning algorithm, can obtain meeting user for the guarantor of the transaction tendency of each insurance Danger transaction portrait, is based on this, the insurance for more meeting customer transaction and being inclined to can be pushed to user, this is going through based on each user What history transaction data was made judges automatically, and compared to the mode of insurance personnel's artificial judgment, can improve and push away to a certain extent Accuracy is recommended, and this workload that automatically processes mode also and can reduce insurance recommendation personnel, improves its working efficiency.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description, The specific work process of device and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
In several embodiments provided by the present invention, it should be understood that disclosed system, device and method can be with It realizes by another way.For example, the apparatus embodiments described above are merely exemplary, for example, the unit It divides, only a kind of logical function partition, there may be another division manner in actual implementation, for example, multiple units or group Part can be combined or can be integrated into another system, or some features can be ignored or not executed.Another point, it is shown Or the mutual coupling, direct-coupling or communication connection discussed can be through some interfaces, device or unit it is indirect Coupling or communication connection can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme 's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list Member both can take the form of hardware realization, can also realize in the form of hardware adds SFU software functional unit.
The above-mentioned integrated unit being realized in the form of SFU software functional unit can store and computer-readable deposit at one In storage media.Above-mentioned SFU software functional unit is stored in a storage medium, including some instructions are used so that a computer It is each that device (can be personal computer, server or network equipment etc.) or processor (Processor) execute the present invention The part steps of embodiment the method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (Read- Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic or disk etc. it is various It can store the medium of program code.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention Within mind and principle, any modification, equivalent substitution, improvement and etc. done be should be included within the scope of the present invention.

Claims (10)

1. a kind of insurance method for pushing based on user's portrait, which is characterized in that the described method includes:
Obtain the historical trading data of user;
Using machine learning algorithm and feature extraction algorithm, the historical trading data is handled, obtains the user's Insurance transaction portrait, the insurance transaction portrait are inclined to degree for the transaction of each insurance to characterize the user;
It is traded and is drawn a portrait according to the insurance, pushed and insure to the user.
2. the method according to claim 1, wherein it is described utilize machine learning algorithm and feature extraction algorithm, The historical trading data is handled, the insurance transaction portrait of the user is obtained, comprising:
The historical trading data is handled using the feature extraction algorithm, the historical trading for obtaining the historical trading data is special Sign and transaction tendency feature;
Using the historical trading feature as input, using the transaction tendency feature as export, using neural network mechanism into Row deep learning obtains insurance transaction probability prediction model;
Each insurance is predicted by the insurance transaction probability prediction model, obtains the transaction that the user is directed to each insurance Tendency degree is drawn a portrait using trading as the insurance.
3. according to the method described in claim 2, it is characterized in that, it is described using the historical trading feature as input, by institute Transaction tendency feature is stated as output, carries out deep learning using neural network mechanism, obtains insurance transaction probability prediction model, Include:
Using the historical trading feature as input, using transaction tendency feature as exporting, using random forests algorithm and At least one of decision tree machine algorithm carries out deep learning, obtains the insurance transaction probability prediction model.
4. method according to any one of claims 1 to 3, which is characterized in that described traded according to the insurance is drawn a portrait, to The user pushes insurance, comprising:
It is traded and is drawn a portrait according to the insurance, determine that target is insured;
Target insurance is pushed to the user.
5. according to the method described in claim 4, it is characterized in that, it is described according to the insurance trade draw a portrait, determine target protect Danger, comprising:
It is directed to the sequence of the transaction tendency degree of each insurance from high to low according to the user, each insurance is ranked up;
The insurance of specifying number from front to back is obtained in the sequence, it is described to specify number to insure as the target For the integer greater than 0.
6. according to the method described in claim 4, it is characterized in that, it is described by the target insurance be pushed to the user, wrap It includes:
Target insurance list is pushed to the user;Alternatively, being sequentially output each target insurance.
7. a kind of insurance driving means based on user's portrait, which is characterized in that described device includes:
Acquiring unit, for obtaining the historical trading data of user;
Processing unit handles the historical trading data, obtains for utilizing machine learning algorithm and feature extraction algorithm To the insurance transaction portrait of the user, the transaction that the insurance transaction portrait is directed to each insurance to characterize the user is inclined to Degree;
Push unit is drawn a portrait for being traded according to the insurance, is pushed and is insured to the user.
8. device according to claim 7, which is characterized in that the processing unit is specifically used for:
The historical trading data is handled using the feature extraction algorithm, the historical trading for obtaining the historical trading data is special Sign and transaction tendency feature;
Using the historical trading feature as input, using the transaction tendency feature as export, using neural network mechanism into Row deep learning obtains insurance transaction probability prediction model;
Each insurance is predicted by the insurance transaction probability prediction model, obtains the transaction that the user is directed to each insurance Tendency degree is drawn a portrait using trading as the insurance.
9. a kind of insurance driving means based on user's portrait characterized by comprising memory, processor and be stored in In the memory and the computer program that can run on the processor, which is characterized in that described in the processor executes It is realized when computer program such as the step of any one of claim 1 to 6 the method.
10. a kind of storage medium characterized by comprising computer executable instructions, when the computer executable instructions quilt When operation, to execute such as the insurance method for pushing as claimed in any one of claims 1 to 6 drawn a portrait based on user.
CN201811013020.7A 2018-08-31 2018-08-31 Insurance method for pushing, device and storage medium based on user's portrait Pending CN109300050A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811013020.7A CN109300050A (en) 2018-08-31 2018-08-31 Insurance method for pushing, device and storage medium based on user's portrait

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811013020.7A CN109300050A (en) 2018-08-31 2018-08-31 Insurance method for pushing, device and storage medium based on user's portrait

Publications (1)

Publication Number Publication Date
CN109300050A true CN109300050A (en) 2019-02-01

Family

ID=65165949

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811013020.7A Pending CN109300050A (en) 2018-08-31 2018-08-31 Insurance method for pushing, device and storage medium based on user's portrait

Country Status (1)

Country Link
CN (1) CN109300050A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111428885A (en) * 2020-03-31 2020-07-17 深圳前海微众银行股份有限公司 User indexing method in federated learning and federated learning device

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111428885A (en) * 2020-03-31 2020-07-17 深圳前海微众银行股份有限公司 User indexing method in federated learning and federated learning device
CN111428885B (en) * 2020-03-31 2021-06-04 深圳前海微众银行股份有限公司 User indexing method in federated learning and federated learning device

Similar Documents

Publication Publication Date Title
Saelens et al. A comparison of single-cell trajectory inference methods
JP6419859B2 (en) Interactive interface for machine learning model evaluation
Chen et al. Entity embedding-based anomaly detection for heterogeneous categorical events
EP1739580B1 (en) Categorization including dependencies between different category systems
EP2579183B1 (en) Biometric training and matching engine
EP2506151A1 (en) Semantic syntax tree kernel-based processing system and method for automatically extracting semantic correlations between scientific and technological core entities
EP1881427A1 (en) Database analysis program, database analysis apparatus, and database analysis method
US9753968B1 (en) Systems and methods for detection of anomalous entities
Xie et al. A combination of boosting and bagging for kdd cup 2009-fast scoring on a large database
Zeng et al. Constructing better classifier ensemble based on weighted accuracy and diversity measure
CN110020660A (en) Use the integrity assessment of the unstructured process of artificial intelligence (AI) technology
CN106776757B (en) Method and device for indicating user to complete online banking operation
CN109300050A (en) Insurance method for pushing, device and storage medium based on user's portrait
KR102038237B1 (en) Credit score model training method, credit score calculation method, apparatus and server
Etz Using Bayes factors to get the most Out of linear regression: A practical guide using R
CN110222171A (en) A kind of application of disaggregated model, disaggregated model training method and device
Parau et al. Using community detection for sentiment analysis
Szafranko Evaluation of data obtained from expert opinions in multi-criteria analyses of construction investment variants
Le Falher et al. On the troll-trust model for edge sign prediction in social networks
Ochodek Approximation of COSMIC functional size of scenario-based requirements in Agile based on syntactic linguistic features—a replication study
CN108009298B (en) Internet character search information integration analysis control method
CN110046196A (en) Identify correlating method and device, electronic equipment
CN109933793A (en) Text polarity identification method, apparatus, equipment and readable storage medium storing program for executing
CN108460630B (en) Method and device for carrying out classification analysis based on user data
Lopez et al. Alternative Ensemble Classifier Based on Penalty Strategy for Improving Prediction Accuracy

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