CN107578270A - A kind of construction method, device and the computing device of financial label - Google Patents
A kind of construction method, device and the computing device of financial label Download PDFInfo
- Publication number
- CN107578270A CN107578270A CN201710655552.XA CN201710655552A CN107578270A CN 107578270 A CN107578270 A CN 107578270A CN 201710655552 A CN201710655552 A CN 201710655552A CN 107578270 A CN107578270 A CN 107578270A
- Authority
- CN
- China
- Prior art keywords
- user
- consumption
- transaction
- matrix
- dimension
- 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
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present embodiments relate to technical field of data processing, more particularly to a kind of construction method, device and the computing device of financial label, the consumption feature of user can not be indicated to solve in the prior art to enter user the method for row label, the problem of accuracy is low.In the embodiment of the present invention, transaction data and user behaviors log with reference to user are obtained；According to the transaction data with reference to user, structure consumption rating matrix, it is the consumption scoring with reference to user on a transaction dimension to consume an element in rating matrix；According to the user behaviors log with reference to user, the vector space model with reference to user is established；For a behavioral data with reference to user, word corresponding to behavioral data is mapped on a transaction dimension in consumption rating matrix with reference to user, scored according to the consumption of behavioral data and the transaction dimension of mapping, it is determined that the comprehensive grading with reference to user；According to the comprehensive grading with reference to user, it is determined that the financial label with reference to user.
Description
Technical field
The present invention relates to technical field of data processing, more particularly to a kind of construction method, device and the calculating of financial label
Equipment.
Background technology
User tag is to take out a label according to information such as the social property of user, habits and customs and consumer behaviors
User's portrait model of change.The model can be widely used in precision marketing, user and industry analysis and personalized service etc.
Field.Current published label construction method is mainly by filtering the web browsing daily record of user, and extraction critical field is as mark
Label mark, and by judging that the label classification belonging to the mark is tagged for user.Occur simultaneously by overall merit label
Time and the information such as the frequency calculate interest-degree of the user under the label, and as the weight of user tag.
Stamp methods of the prior art only using the user behaviors log of user as the foundation for determining label, i.e., are existed based on user
Navigation patterns on internet enter row label to user.The label construction method based on navigation patterns this first can produce largely
Semantic redundancy text label, this causes to be difficult in label system from magnanimity, with information redundancy during the label
Targetedly label is found out to be used to analyze.Further, since the particularity of finance and payment industry, many commonly-used label can not
Directly obtained from the user behaviors log of user.Therefore, the user behaviors log according only to user enters the method for row label and can not often marked
The consumption feature of user is shown, accuracy is relatively low.
The content of the invention
The application provides a kind of construction method, device and the computing device of financial label, right in the prior art to solve
The problem of method that user enters row label can not indicate the consumption feature of user, and accuracy is low.
A kind of construction method of financial label provided in an embodiment of the present invention, including：
Obtain the transaction data and user behaviors log with reference to user；
According to the transaction data with reference to user, structure consumption rating matrix, one consumed in rating matrix
Element is the consumption scoring with reference to user on a transaction dimension；
According to the user behaviors log with reference to user, the vector space model with reference to user is established, the vector is empty
Between model include multiple behavioral datas with reference to user, each behavioral data corresponds to the user behaviors log with reference to user
In a word；
For a behavioral data with reference to user, word corresponding to the behavioral data is mapped to the consumption
On a transaction dimension described in rating matrix with reference to user, according to the consumption of the behavioral data and the transaction dimension of mapping
Scoring, determines the comprehensive grading with reference to user；
According to the comprehensive grading with reference to user, the financial label with reference to user is determined.
Optionally, described according to all transaction data with reference to user, structure consumes rating matrix, including：
User is referred to for each, using the transaction data with reference to user, calculates the reference user in different friendships
The condition of consumption of easy dimension；According to the condition of consumption, calculating is described to score with reference to user in the consumption of each transaction dimension；
Scored using all with reference to user in the consumption of each transaction dimension, structure consumption rating matrix.
Optionally, scored using below equation calculating is described with reference to user in the consumption of a transaction dimension：
Wherein, Score is consumption scoring of the user in a transaction dimension, and θ is the weight of the transaction dimension；ω
For it is described with reference to user the transaction dimension consumption stroke count and spending amount weighted average；υ refers to user to be all
In the consumption average of the transaction dimension, σ be it is all with reference to user the transaction dimension variance；User is referred to be described
In the spending amount and the ratio of all spending amount sums with reference to user of the transaction dimension.
Optionally, it is described according to all transaction data with reference to user, after structure consumption rating matrix, in addition to：
Using the method for matrix decomposition, completion is carried out to the incomplete value in the consumption rating matrix.
Optionally, the method using matrix decomposition, completion, bag are carried out to the incomplete value in the consumption rating matrix
Include：
First parameter line vector sum the second parameter row vector of random generation, the element number of the first parameter row vector with
The line number of the consumption rating matrix is equal, the row of the element number of the second parameter row vector and the consumption rating matrix
Number is equal；
The second parameter row vector according to the first parameter line vector sum, calculate the mistake of the consumption rating matrix
Difference；
Second parameter row vector described in the first parameter line vector sum according to the error update, and repeat step according to
Second parameter row vector described in the first parameter line vector sum, the error of the consumption rating matrix is calculated, until the mistake
Difference convergence；
The second parameter row vector determines the consumption rating matrix after completion according to the first parameter line vector sum.
Optionally, it is described that word corresponding to the behavioral data is mapped to described in the consumption rating matrix with reference to use
On one transaction dimension at family, including：
For a transaction dimension with reference to user, transaction dimension word corresponding with the behavioral data is calculated
The similarity of language；
From the All Activity dimension with reference to user, it is determined that the similarity of word corresponding with the behavioral data is most
High transaction dimension；
Word corresponding to the behavioral data is mapped on the similarity highest transaction dimension.
Optionally, the quantity with reference to user is multiple；
It is described that the financial label with reference to user is determined according to the comprehensive grading with reference to user, including：
According to business rule and the background information with reference to user, belong to same class label with reference to determination in user from all
With reference to user；
According to the comprehensive grading for the reference user for belonging to same class label, the forecast model of such label is determined；
According to the forecast model of all kinds of labels, obtain integrating labeling model；
According to the comprehensive grading with reference to user and the comprehensive labeling model, the gold with reference to user is determined
Melt label.
Optionally, the consumption according to the behavioral data and the transaction dimension of mapping is scored, and is determined described with reference to use
After the comprehensive grading at family, in addition to：
It is determined that forming the historical time of comprehensive grading matrix, the comprehensive grading matrix is that all synthesis with reference to user are commented
It is grouped into；
According to the historical time and current time, calculate under the current time, the comprehensive grading matrix after decay；
Comprehensive grading matrix after the decay is calculated according to below equation,
Wherein, α is decay factor, and t is current time, and T is historical time, and M (T) is the comprehensive grading square under historical time
Battle array, M (t) are the comprehensive grading matrix under current time, and M ' (t) is the comprehensive grading matrix after the decay.
A kind of construction device of financial label, including：
Acquiring unit, for obtaining transaction data and user behaviors log with reference to user；
Transaction handling unit, for according to the transaction data with reference to user, structure consumption rating matrix, the consumption
An element in rating matrix is the consumption scoring with reference to user on a transaction dimension；
Text-processing unit, for according to the user behaviors log with reference to user, it is empty to establish the vector with reference to user
Between model, the vector space model includes multiple behavioral datas with reference to user, described in each behavioral data is corresponding
With reference to user user behaviors log in a word；
Computing unit is combined, for for a behavioral data with reference to user, by corresponding to the behavioral data
Word is mapped on a transaction dimension described in the consumption rating matrix with reference to user, according to the behavioral data and is reflected
The consumption scoring for the transaction dimension penetrated, determines the comprehensive grading with reference to user；
Tag unit, for according to the comprehensive grading with reference to user, determining the financial label with reference to user.
Optionally, the transaction handling unit, is specifically used for：
User is referred to for each, using the transaction data with reference to user, calculates the reference user in different friendships
The condition of consumption of easy dimension；According to the condition of consumption, calculating is described to score with reference to user in the consumption of each transaction dimension；
Scored using all with reference to user in the consumption of each transaction dimension, structure consumption rating matrix.
Optionally, the transaction handling unit, specifically for calculating the reference user in a friendship using below equation
The consumption scoring of easy dimension：
Wherein, Score is consumption scoring of the user in a transaction dimension, and θ is the weight of the transaction dimension；ω
For it is described with reference to user the transaction dimension consumption stroke count and spending amount weighted average；υ refers to user to be all
In the consumption average of the transaction dimension, σ be it is all with reference to user the transaction dimension variance；User is referred to be described
In the spending amount and the ratio of all spending amount sums with reference to user of the transaction dimension.
Optionally, the transaction handling unit, is additionally operable to：
Using the method for matrix decomposition, completion is carried out to the incomplete value in the consumption rating matrix.
Optionally, the transaction handling unit, is specifically used for：
First parameter line vector sum the second parameter row vector of random generation, the element number of the first parameter row vector with
The line number of the consumption rating matrix is equal, the row of the element number of the second parameter row vector and the consumption rating matrix
Number is equal；
The second parameter row vector according to the first parameter line vector sum, calculate the mistake of the consumption rating matrix
Difference；
Second parameter row vector described in the first parameter line vector sum according to the error update, and repeat step according to
Second parameter row vector described in the first parameter line vector sum, the error of the consumption rating matrix is calculated, until the mistake
Difference convergence；
The second parameter row vector determines the consumption rating matrix after completion according to the first parameter line vector sum.
Optionally, the combination computing unit, is specifically used for：
For a transaction dimension with reference to user, transaction dimension word corresponding with the behavioral data is calculated
The similarity of language；
From the All Activity dimension with reference to user, it is determined that the similarity of word corresponding with the behavioral data is most
High transaction dimension；
Word corresponding to the behavioral data is mapped on the similarity highest transaction dimension.
Optionally, the quantity with reference to user is multiple；
The tag unit, is specifically used for：
According to business rule and the background information with reference to user, belong to same class label with reference to determination in user from all
With reference to user；
According to the comprehensive grading for the reference user for belonging to same class label, the forecast model of such label is determined；
According to the forecast model of all kinds of labels, obtain integrating labeling model；
According to the comprehensive grading with reference to user and the comprehensive labeling model, the gold with reference to user is determined
Melt label.
Optionally, the combination computing unit, is additionally operable to：
It is determined that forming the historical time of comprehensive grading matrix, the comprehensive grading matrix is that all synthesis with reference to user are commented
It is grouped into；
According to the historical time and current time, calculate under the current time, the comprehensive grading matrix after decay；
Comprehensive grading matrix after the decay is calculated according to below equation,
Wherein, α is decay factor, and t is current time, and T is historical time, and M (T) is the comprehensive grading square under historical time
Battle array, M (t) are the comprehensive grading matrix under current time, and M ' (t) is the comprehensive grading matrix after the decay.
A kind of computing device, including：
Memory, instructed for storage program；
Processor, for calling the programmed instruction stored in the memory, performed according to the program of acquisition：Obtain reference
The transaction data and user behaviors log of user；It is described to disappear according to the transaction data with reference to user, structure consumption rating matrix
It is the consumption scoring with reference to user on a transaction dimension to take an element in rating matrix；According to described with reference to use
The user behaviors log at family, establishes the vector space model with reference to user, and the vector space model includes described with reference to use
Multiple behavioral datas at family, each behavioral data correspond to a word in the user behaviors log with reference to user；For described
With reference to the behavioral data of user, word corresponding to the behavioral data is mapped to described in the consumption rating matrix and joined
Examine on the transaction dimension of user, scored according to the consumption of the behavioral data and the transaction dimension of mapping, determine the ginseng
Examine the comprehensive grading of user；According to the comprehensive grading with reference to user, the financial label with reference to user is determined.
In the embodiment of the present invention, optional part user obtains the friendship with reference to user as user is referred to from all users
Easy data and user behaviors log.According to the transaction data with reference to user, structure consumes rating matrix, wherein, one in consumption scoring
Individual element is the consumption scoring on a transaction dimension with reference to user.Meanwhile according to the user behaviors log with reference to user, establish ginseng
The vector space model of user is examined, the vector space model includes multiple behavioral datas with reference to user, each behavioral data
A word in the corresponding user behaviors log with reference to user.The embodiment of the present invention is not only used as label structure using the user behaviors log of user
The foundation built, the transaction data of user is reference is also made to, and both are merged.Specifically, for one with reference to user
Behavioral data, word corresponding to the behavioral data is mapped to described in the consumption rating matrix with reference to the friendship of user
In easy dimension, and scored according to the consumption of behavioral data and the transaction dimension of mapping, determine the comprehensive grading for referring to user.Most
Afterwards, according to the comprehensive grading, it is determined that the comprehensive grading with reference to user.The embodiment of the present invention is liked and handed over by the behavior to user
Easy situation carries out overall merit, establishes the label of energy accurate description user's finance preference and consumption feature, solves according to user
User behaviors log enter the problem of method of row label can not often indicate the consumption feature of user.And compared with prior art,
The embodiment of the present invention builds the feature of user, due to disappearing for all users by being merged to user behaviors log and transaction data
Take score data and integrate decision by the transaction details and user behaviors log of user, this mode is reduced in conventional method according to meter
Cumulative errors caused by similarity between the text of calculation user behaviors log, add the degree of accuracy of label foundation.
Brief description of the drawings
Technical scheme in order to illustrate the embodiments of the present invention more clearly, make required in being described below to embodiment
Accompanying drawing is briefly introduced, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for this
For the those of ordinary skill in field, without having to pay creative labor, it can also be obtained according to these accompanying drawings
His accompanying drawing.
Fig. 1 is a kind of schematic flow sheet of the construction method of financial label provided in an embodiment of the present invention；
Fig. 2 is the schematic flow sheet for the method that label is established in the specific embodiment of the invention；
Fig. 3 is a kind of structural representation of the construction device of financial label provided in an embodiment of the present invention.
Embodiment
In order that the object, technical solutions and advantages of the present invention are clearer, the present invention is made below in conjunction with accompanying drawing into
One step it is described in detail, it is clear that the described embodiment only a part of embodiment of the present invention, rather than whole implementation
Example.Based on the embodiment in the present invention, what those of ordinary skill in the art were obtained under the premise of creative work is not made
All other embodiment, belongs to the scope of protection of the invention.
Fig. 1 illustrates a kind of schematic flow sheet of the construction method of financial label provided in an embodiment of the present invention.
As shown in figure 1, the construction method of financial label provided in an embodiment of the present invention, comprises the following steps：
Step 101, obtain the transaction data and user behaviors log for referring to user.
Step 102, according to the transaction data with reference to user, structure consumption rating matrix, the consumption rating matrix
In an element for it is described with reference to user on a transaction dimension consumption scoring.
Step 103, according to the user behaviors log with reference to user, establish the vector space model with reference to user, institute
Stating vector space model includes multiple behavioral datas with reference to user, and each behavioral data corresponds to described with reference to user's
A word in user behaviors log.
Step 104, for a behavioral data with reference to user, word corresponding to the behavioral data is mapped to
On a transaction dimension described in the consumption rating matrix with reference to user, tieed up according to the transaction of the behavioral data and mapping
The consumption scoring of degree, determines the comprehensive grading with reference to user.
Step 105, according to the comprehensive grading with reference to user, determine the financial label with reference to user.
In the embodiment of the present invention, optional part user obtains the friendship with reference to user as user is referred to from all users
Easy data and user behaviors log.According to the transaction data with reference to user, structure consumes rating matrix, wherein, one in consumption scoring
Individual element is the consumption scoring on a transaction dimension with reference to user.Meanwhile according to the user behaviors log with reference to user, establish ginseng
The vector space model of user is examined, the vector space model includes multiple behavioral datas with reference to user, each behavioral data
A word in the corresponding user behaviors log with reference to user.The embodiment of the present invention is not only used as label structure using the user behaviors log of user
The foundation built, the transaction data of user is reference is also made to, and both are merged.Specifically, for one with reference to user
Behavioral data, word corresponding to the behavioral data is mapped to described in the consumption rating matrix with reference to the friendship of user
In easy dimension, and scored according to the consumption of behavioral data and the transaction dimension of mapping, determine the comprehensive grading for referring to user.Most
Afterwards, according to the comprehensive grading, it is determined that the comprehensive grading with reference to user.The embodiment of the present invention is liked and handed over by the behavior to user
Easy situation carries out overall merit, establishes the label of energy accurate description user's finance preference and consumption feature, solves according to user
User behaviors log enter the problem of method of row label can not often indicate the consumption feature of user.And compared with prior art,
The embodiment of the present invention builds the feature of user, due to disappearing for all users by being merged to user behaviors log and transaction data
Take score data and integrate decision by the transaction details and user behaviors log of user, this mode is reduced in conventional method according to meter
Cumulative errors caused by similarity between the text of calculation user behaviors log, add the degree of accuracy of label foundation.
In the embodiment of the present invention, for reference to user establish label be based on reference to user transaction data and user behaviors log,
Therefore, it is necessary first to obtain the transaction data with reference to user, and handled.
Above-mentioned steps 102, described according to all transaction data with reference to user, structure consumes rating matrix, including：
User is referred to for each, using the transaction data with reference to user, calculates the reference user in different friendships
The condition of consumption of easy dimension；According to the condition of consumption, calculating is described to score with reference to user in the consumption of each transaction dimension；
Scored using all with reference to user in the consumption of each transaction dimension, structure consumption rating matrix.
Specifically, the transaction data with reference to user is obtained, and the consumption for each referring to user is calculated according to transaction data
Situation, the condition of consumption can be with reference to the consumption stroke count of user or spending amount etc..In order that calculate more targeted, the present invention
In embodiment according to transaction data statistical-reference user different transaction dimensions the condition of consumption.Specially 5 transaction dimensions, point
It is not transaction geography, time bracket, dealing money section, transaction channel and transaction merchant type.Afterwards, according to reference to user's
The condition of consumption calculates it and scored in the consumption of each transaction dimension, can profit in the consumption scoring of a transaction dimension with reference to user
Calculated with below equation：
Wherein, Score is consumption scoring of the user in a transaction dimension, and θ is the weight of the transaction dimension；ω
For it is described with reference to user the transaction dimension consumption stroke count and spending amount weighted average；υ refers to user to be all
In the consumption average of the transaction dimension, σ be it is all with reference to user the transaction dimension consumption variance；For the reference
User is in the spending amount of the transaction dimension and the ratio of all spending amount sums with reference to user.
For example, it is the consumption scoring that merchant type is cosmetics of merchandising to calculate user a in time bracket.By formula 1, θ
For the weight of cosmetics, the value can determine according to the service conditions of corresponding transaction dimension.ω is user a on cosmetics
Consume the weighted average of stroke count and spending amount.υ is all consumption averages with reference to user on cosmetics.σ is all ginsengs
Examine consumption variance of the user on cosmetics.For spending amounts of the user a on cosmetics and user a all spending amounts
The ratio of sum.
Using formula 1, one can be calculated and scored with reference to user in the consumption of each transaction dimension, and according to all ginsengs
Examine the consumption scoring structure consumption rating matrix of user.
Because part is relatively low with reference to the trading activity frequency of user, causes these to have with reference to user in many dimensions and do not merchandise
Situation, the incomplete values of a large amount of consumption scorings can be formed.Therefore, step 102, it is described according to all numbers of deals with reference to user
According to, after structure consumption rating matrix, in addition to：
Using the method for matrix decomposition, completion is carried out to the incomplete value in the consumption rating matrix.
Specifically, the method using matrix decomposition, completion is carried out to the incomplete value in the consumption rating matrix,
Including：
First parameter line vector sum the second parameter row vector of random generation, the element number of the first parameter row vector with
The line number of the consumption rating matrix is equal, the row of the element number of the second parameter row vector and the consumption rating matrix
Number is equal；
The second parameter row vector according to the first parameter line vector sum, calculate the mistake of the consumption rating matrix
Difference；
Second parameter row vector described in the first parameter line vector sum according to the error update, and repeat step according to
Second parameter row vector described in the first parameter line vector sum, the error of the consumption rating matrix is calculated, until the mistake
Difference convergence；
The second parameter row vector determines the consumption rating matrix after completion according to the first parameter line vector sum.
Generally we use rating matrix MR×NRepresent all to score in the consumption of each dimension with reference to user, incompleteness value is by 0 generation
Replace.Wherein 1≤r≤R represents user index, i.e., shared R refer to user, and 1≤n≤N represents transaction dimension index, i.e., each to use
There is N number of transaction dimension at family.ψ (r) represents customer parameter vector, and ψ (n) is dimensional parameter vector.It is desirable that find suitable ginseng
Number vector ψ (r) and ψ (n) causes MR×N=ψ (r)T·ψ(n).Specific method is as follows：
Step 1: input rating matrix MR×N, and random initializtion generation customer parameter vector ψ (r) and dimensional parameter vector
ψ(n).Wherein, customer parameter vector is the first parameter vector, and dimensional parameter vector is the second parameter vector, due to customer parameter
The element number of vector is equal to the transaction dimension number for referring to user, therefore the element number of customer parameter vector and scoring square
Battle array MR×NLine number it is equal.Accordingly, the element number and rating matrix M of dimensional parameter vectorR×NColumns it is equal.
Step 2: calculate rating matrix MR×NMiddle nonzero element and ψ (r)Tψ (n) error, that is, calculate εR×N=MR×N-ψ
(r)Tψ (n) ... formula 2.
Step 3: according to error update parameter vector ψ (r) and ψ (n), calculation formula is：
ψ (r)=ψ (r)+α [εR×N·ψ(n)-λrψ (r)] ... formula 3,
ψ (n)=ψ (n)+α [εR×N·ψ(r)-λnψ (n)] ... formula 4.
Wherein, α and λ is study renewal rate, and α is overall rate, and its numerical value is usually 0.05.Because customer parameter to
It is often a longer vector to measure ψ (r), and dimensional parameter vector ψ (n) is often shorter, in order to distinguish two vectorial renewal speed
Rate, introduce parameter lambda.Wherein λrFor user vector ψ (r) renewal rate, typically smaller than α, λnFor user vector ψ (r) renewal speed
Rate, typically larger than α.
Step 4: repeat step two and step 3, until error matrix εR×NIt is stable, error matrix εR×NIt is stable to refer to
The average of the absolute value of the difference of the secondary each element for calculating errors matrix and error current matrix is less than some fixed value.
Step 5: the complete rating matrix after output completion
In the embodiment of the present invention, except the foundation built using the transaction data with reference to user as label and handled,
It reference is also made to the transaction data of user.Further, step 104, word corresponding to the behavioral data is mapped to the consumption
On a transaction dimension described in rating matrix with reference to user, including：
For a transaction dimension with reference to user, transaction dimension word corresponding with the behavioral data is calculated
The similarity of language；
From the All Activity dimension with reference to user, it is determined that the similarity of word corresponding with the behavioral data is most
High transaction dimension；
Word corresponding to the behavioral data is mapped on the similarity highest transaction dimension.
In particular it is necessary to handle the behavioral data with reference to user, processing mode here is directed to existing
The processing method of the text data of user is similar.To same the log informations such as information, comment, user behaviors log are browsed with reference to user
It is collected, cleans and filters, this is aggregated into this with reference to all log informations of user refers to document corresponding to user, utilizes
TF-IDF (term frequency-inverse document frequency, the conventional weighting skill of information retrieval data mining
Art) the text vector spatial model for referring to user is established, text vector spatial model includes each row for referring to user
For word and corresponding behavioral data.
Afterwards, the behavioral data with reference to user is merged with transaction dimension.It will be joined by the method for calculating similarity
The behavioral data for examining user is mapped on transaction dimension, and Hownet and news category corpus can be utilized to calculate and refer to user behavior
The degree of correlation of word and transaction dimension, similarity that specifically can be by similarity that corpus learns out with being calculated based on Hownet
It is combined as last similarity.It can utilize ranking operation that news category is mapped into transaction dimension, wherein weighting
Weight for transaction dimension and behavior word text similarity.
For example, it is initial on transaction dimension " food and drink " with reference to user to calculate this by reference to the transaction data of user
Consumption scoring is ScoreFood and drink=0.1, then it is this is right with reference to all behavior words (such as " grilled fish ") similar to " food and drink " of user
In the consumption scoring of food and drink, specific formula can be for the behavioral data weighting answered：
ScoreAlways=ScoreFood and drink+s×ScoreGrilled fish... formula 6
Wherein, ScoreGrilled fishFor the scoring of " grilled fish " that is calculated according to TFIDF algorithms, s is the text of " grilled fish " and " food and drink "
Similarity.
It is a kind of time series data in view of the condition of consumption for referring to user, understands change over time and change, therefore, needs
Comprehensive grading with reference to user is decayed.The consumption according to the behavioral data and the transaction dimension of mapping is commented
Point, after determining the comprehensive grading with reference to user, in addition to：
It is determined that forming the historical time of comprehensive grading matrix, the comprehensive grading matrix is that all synthesis with reference to user are commented
It is grouped into；
According to the historical time and current time, calculate under the current time, the comprehensive grading matrix after decay；
Comprehensive grading matrix after the decay is calculated according to below equation,
Wherein, α is decay factor, and t is current time, and T is historical time, and M (T) is the comprehensive grading square under historical time
Battle array, M (t) are the comprehensive grading matrix under current time, and M ' (t) is the comprehensive grading matrix after the decay.
Finally, it is that user establishes label according to comprehensive grading matrix, mainly considers how to establish grader.Due to establishing
Label is a kind of statistics, so the quantity for establishing the reference user of grader institute foundation is multiple, herein with reference to user's
Quantity is more, and the grader finally established is more accurate.
It is described that the financial label with reference to user is determined according to the comprehensive grading with reference to user, including：
According to business rule and the background information with reference to user, belong to same class label with reference to determination in user from all
With reference to user；
According to the comprehensive grading for the reference user for belonging to same class label, the forecast model of such label is determined；
According to the forecast model of all kinds of labels, obtain integrating labeling model；
According to the comprehensive grading with reference to user and the comprehensive labeling model, the gold with reference to user is determined
Melt label.
Specifically, according to business rule and the background information with reference to user, all reference users are established into label.Example
Such as, it can be the label system established with reference to user, be broadly divided into following four major class：(1) ascribed characteristics of population：Such as sex, the age,
Level of consumption etc..(2) status attribute：Such as whether card amount needs to be lifted, if has room, if having car, if has child etc..
(3) transaction attribute：For example whether often input password by mistake, if often Sorry, your ticket has not enough value.(4) finance and consumption preferences：Such as manage
Wealth, insurance, number, enchashment fan, Yun Shanfu intelligent etc..For the ascribed characteristics of population and the class label of status attribute two, we pass through
Professional knowledge and the objective background data of reference user are built jointly.By taking the sex label in the ascribed characteristics of population as an example, 50% number
According to determined by business rule, if such as with reference to user the dimension such as " tobacco and wine ", " men's clothing " consume it is more frequent, then it is assumed that the ginseng
It is male to examine user.If existing with reference to user " cosmetics ", the dimension such as " beauty " consumption it is more frequent, then it is assumed that be women.Training set
In 50% data by with reference to user background information in obtain.For consumption preferences class label, we are by verifying with reference to use
Whether family was used as checking in one month in dimension consumption is gathered.
According to the label with reference to user, all from comprehensive grading matrix multiple refer to user, work with reference to selecting in users
Added for training sample in training set, and the power of each training sample in the training set is determined according to the quantity of training sample
Value；It is k groups by all training sample random divisions, using the training sample of maximum weight in every group as test sample；According to institute
State test sample and determine grader；All training samples are classified using the grader, and with the label of training sample
Compare, determine the error rate and grader weight of the grader；According to the grader weight to all training samples
Weights be updated, and all training sample random divisions are k groups by repeat step, until the grader number determined is more than
Threshold value；All graders are subjected to linear weighted function, determine integrated classifier；Using the integrated classifier user is referred to all
Classified, added part in the training set as training sample with reference to user according to classification results, and repeat step root
The weights of each training sample in the training set are determined according to the quantity of training sample, until the training sample in the training set
No longer change；Using the integrated classifier that training sample no longer changes as financial grader, gold is established for the user to be sorted
Melt label.
The above-mentioned flow for establishing label is described in detail with specific embodiment below, as shown in Fig. 2 specific steps bag
Include：
Step 201, initialization.N is selected from all reference users as training sample to add in training set T, it is assumed that
Training set T={ (x1,y1),(x2,y2),……(xn,yn), wherein, ynUser, x are referred to for n-thnUser is referred to for n-th
The corresponding vector in comprehensive grading matrix.The weights of each training sample in training set are initialized, all weight W are entered asI.e.
Step 202, stochastical sampling structure grader.It is k groups by all training sample random divisions, by weights W in k groups most
Big training sample trains test sample to obtain grader G as test samplei。
Step 203, weight renewal.Utilize grader GiEach training sample in training set T is classified.Will be each
The classification of individual training sample and the label of the training sample compare, and calculate the error of each training sample, so as to statistical
Class device GiError rate ei, and grader G is calculated according to following equationiWeight
Wherein as error rate eiWhen≤0.5, grader weightThat is its smaller weight of grader error rate is bigger.
Whether step 204, the grader number for judging to determine are more than threshold value, if so, performing step 206；Otherwise step is performed
Rapid 205.
Step 205, using following formula training sample weight is updated, performs step 202 afterwards.
Wherein δjFor two-valued function, work as Giδ when correct to training sample x classificationjFor 1, otherwise δjFor 0.
Step 206, by all graders carry out linear weighted function, obtain integrated classifier
Step 207, classified to all with reference to users using the integrated classifier that step 205 obtains, tied according to classification
Fruit adds score higher reference user in training set.
Whether the training sample that step 208, training of judgement are concentrated changes, if so, then performing step 201, otherwise performs step
Rapid 209.
Step 209, using the integrated classifier that training sample no longer changes as financial grader, be using financial grader
User to be sorted establishes financial label, and the weight using the accuracy rate of financial grader as label.
The embodiment of the present invention is merged by the behavioral data to user and transaction data, builds intermediate layer characteristic
According to being learnt using machine learning method to user tag, obtain the confidence level of user tag.Compared with prior art, originally
All transaction data with reference to user are calculated by with reference to the transaction details of user in invention, reduce conventional method due to
Cumulative errors caused by similarity between calculating text so that the label ultimately generated has the higher degree of accuracy.Meanwhile relative to
The method of the frequent field generation label of traditional basis, the embodiment of the present invention are directly established label system by business rule, marked
Its correctness is can verify that during label generation, so, the label of generation has more specific aim.In addition, the embodiment of the present invention is kept away
Exempt from mode of the conventional method using label frequency of occurrence as label weight, but by way of constructing training data, utilize
Machine learning algorithm calculates label confidence level, and using this as label weight.This greatly facilitates the follow-up use of label, label
User can be by selecting corresponding confidence level to obtain target group.
Fig. 3 illustrates a kind of structural representation of the construction device of financial label provided in an embodiment of the present invention.
As shown in figure 3, a kind of construction device of financial label provided in an embodiment of the present invention, including：
Acquiring unit 301, for obtaining transaction data and user behaviors log with reference to user；
Transaction handling unit 302, for according to the transaction data with reference to user, structure consumption rating matrix
It is the consumption scoring with reference to user on a transaction dimension to consume an element in rating matrix；
Text-processing unit 303, for according to the user behaviors log with reference to user, establishing the vector with reference to user
Spatial model, the vector space model include multiple behavioral datas with reference to user, and each behavioral data corresponds to institute
State a word in the user behaviors log with reference to user；
Computing unit 304 is combined, for for a behavioral data with reference to user, the behavioral data to be corresponded to
Word be mapped on a transaction dimension described in the consumption rating matrix with reference to user, according to the behavioral data and
The consumption scoring of the transaction dimension of mapping, determines the comprehensive grading with reference to user；
Tag unit 305, for according to the comprehensive grading with reference to user, determining the finance mark with reference to user
Label.
Optionally, the transaction handling unit 302, is specifically used for：
User is referred to for each, using the transaction data with reference to user, calculates the reference user in different friendships
The condition of consumption of easy dimension；According to the condition of consumption, calculating is described to score with reference to user in the consumption of each transaction dimension；
Scored using all with reference to user in the consumption of each transaction dimension, structure consumption rating matrix.
Optionally, the transaction handling unit 302, specifically for calculating the reference user at one using below equation
The consumption scoring of transaction dimension：
Wherein, Score is consumption scoring of the user in a transaction dimension, and θ is the weight of the transaction dimension；ω
For it is described with reference to user the transaction dimension consumption stroke count and spending amount weighted average；υ refers to user to be all
In the consumption average of the transaction dimension, σ be it is all with reference to user the transaction dimension variance；User is referred to be described
In the spending amount and the ratio of all spending amount sums with reference to user of the transaction dimension.
Optionally, the transaction handling unit 302, is additionally operable to：
Using the method for matrix decomposition, completion is carried out to the incomplete value in the consumption rating matrix.
Optionally, the transaction handling unit, is specifically used for：
First parameter line vector sum the second parameter row vector of random generation, the element number of the first parameter row vector with
The line number of the consumption rating matrix is equal, the row of the element number of the second parameter row vector and the consumption rating matrix
Number is equal；
The second parameter row vector according to the first parameter line vector sum, calculate the mistake of the consumption rating matrix
Difference；
Second parameter row vector described in the first parameter line vector sum according to the error update, and repeat step according to
Second parameter row vector described in the first parameter line vector sum, the error of the consumption rating matrix is calculated, until the mistake
Difference convergence；
The second parameter row vector determines the consumption rating matrix after completion according to the first parameter line vector sum.
Optionally, the combination computing unit 304, is specifically used for：
For a transaction dimension with reference to user, transaction dimension word corresponding with the behavioral data is calculated
The similarity of language；
From the All Activity dimension with reference to user, it is determined that the similarity of word corresponding with the behavioral data is most
High transaction dimension；
Word corresponding to the behavioral data is mapped on the similarity highest transaction dimension.
Optionally, the quantity with reference to user is multiple；
The tag unit 305, is specifically used for：
According to business rule and the background information with reference to user, belong to same class label with reference to determination in user from all
With reference to user；
According to the comprehensive grading for the reference user for belonging to same class label, the forecast model of such label is determined；
According to the forecast model of all kinds of labels, obtain integrating labeling model；
According to the comprehensive grading with reference to user and the comprehensive labeling model, the gold with reference to user is determined
Melt label.
Optionally, the combination computing unit 304, is additionally operable to：
It is determined that forming the historical time of comprehensive grading matrix, the comprehensive grading matrix is that all synthesis with reference to user are commented
It is grouped into；
According to the historical time and current time, calculate under the current time, the comprehensive grading matrix after decay；
Comprehensive grading matrix after the decay is calculated according to below equation,
Wherein, α is decay factor, and t is current time, and T is historical time, and M (T) is the comprehensive grading square under historical time
Battle array, M (t) are the comprehensive grading matrix under current time, and M ' (t) is the comprehensive grading matrix after the decay.
The embodiments of the invention provide a kind of computing device, the computing device is specifically as follows desktop computer, portable
Computer, smart mobile phone, tablet personal computer, personal digital assistant (Personal Digital Assistant, PDA) etc..The meter
Central processing unit (Center Processing Unit, CPU), memory, input-output apparatus etc. can be included by calculating equipment,
Input equipment can include keyboard, mouse, touch-screen etc., and output equipment can include display device, such as liquid crystal display
(Liquid Crystal Display, LCD), cathode-ray tube (Cathode Ray Tube, CRT) etc..
Memory can include read-only storage (ROM) and random access memory (RAM), and provide storage to processor
The programmed instruction and data stored in device.In embodiments of the present invention, memory can be used for the construction method of financial label
Program.
By calling the programmed instruction of memory storage, processor is used to perform according to the programmed instruction of acquisition processor：
Obtain the transaction data and user behaviors log with reference to user；According to the transaction data with reference to user, structure consumption scoring square
Battle array, the element consumed in rating matrix are the consumption scoring with reference to user on a transaction dimension；According to
The user behaviors log with reference to user, establishes the vector space model with reference to user, and the vector space model includes
Multiple behavioral datas with reference to user, each behavioral data correspond to a word in the user behaviors log with reference to user
Language；For a behavioral data with reference to user, word corresponding to the behavioral data is mapped to the consumption and scored
On a transaction dimension described in matrix with reference to user, commented according to the consumption of the behavioral data and the transaction dimension of mapping
Point, determine the comprehensive grading with reference to user；According to the comprehensive grading with reference to user, the gold with reference to user is determined
Melt label.
The present invention is the flow with reference to method according to embodiments of the present invention, equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that can be by every first-class in computer program instructions implementation process figure and/or block diagram
Journey and/or the flow in square frame and flow chart and/or block diagram and/or the combination of square frame.These computer programs can be provided
The processors of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing devices is instructed to produce
A raw machine so that produced by the instruction of computer or the computing device of other programmable data processing devices for real
The device for the function of being specified in present one flow of flow chart or one square frame of multiple flows and/or block diagram or multiple square frames.
These computer program instructions, which may be alternatively stored in, can guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works so that the instruction being stored in the computer-readable memory, which produces, to be included referring to
Make the manufacture of device, the command device realize in one flow of flow chart or multiple flows and/or one square frame of block diagram or
The function of being specified in multiple square frames.
These computer program instructions can be also loaded into computer or other programmable data processing devices so that counted
Series of operation steps is performed on calculation machine or other programmable devices to produce computer implemented processing, so as in computer or
The instruction performed on other programmable devices is provided for realizing in one flow of flow chart or multiple flows and/or block diagram one
The step of function of being specified in individual square frame or multiple square frames.
Although preferred embodiments of the present invention have been described, but those skilled in the art once know basic creation
Property concept, then can make other change and modification to these embodiments.So appended claims be intended to be construed to include it is excellent
Select embodiment and fall into having altered and changing for the scope of the invention.
Obviously, those skilled in the art can carry out the essence of various changes and modification without departing from the present invention to the present invention
God and scope.So, if these modifications and variations of the present invention belong to the scope of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to including these changes and modification.
Claims (17)
- A kind of 1. construction method of financial label, it is characterised in that including：Obtain the transaction data and user behaviors log with reference to user；According to the transaction data with reference to user, structure consumption rating matrix, the element consumed in rating matrix For the consumption scoring with reference to user on a transaction dimension；According to the user behaviors log with reference to user, the vector space model with reference to user, the vector space mould are established Type includes multiple behavioral datas with reference to user, and each behavioral data is corresponded in the user behaviors log with reference to user One word；For a behavioral data with reference to user, word corresponding to the behavioral data is mapped to the consumption and scored On a transaction dimension described in matrix with reference to user, commented according to the consumption of the behavioral data and the transaction dimension of mapping Point, determine the comprehensive grading with reference to user；According to the comprehensive grading with reference to user, the financial label with reference to user is determined.
- 2. the method as described in claim 1, it is characterised in that described to be disappeared according to all transaction data with reference to user, structure Take rating matrix, including：User is referred to for each, using the transaction data with reference to user, calculates the reference user in different transaction dimensions The condition of consumption of degree；According to the condition of consumption, calculating is described to score with reference to user in the consumption of each transaction dimension；Scored using all with reference to user in the consumption of each transaction dimension, structure consumption rating matrix.
- 3. method as claimed in claim 2, it is characterised in that calculate the reference user in a transaction using below equation The consumption scoring of dimension：Wherein, Score is consumption scoring of the user in a transaction dimension, and θ is the weight of the transaction dimension；ω is institute State with reference to user in the consumption stroke count of the transaction dimension and the weighted average of spending amount；υ is all reference users in institute State the consumption average of transaction dimension, σ be it is all with reference to user the transaction dimension variance；Be it is described with reference to user in institute State the spending amount of transaction dimension and the ratio of all spending amount sums with reference to user.
- 4. the method as described in claim 1, it is characterised in that described to be disappeared according to all transaction data with reference to user, structure After taking rating matrix, in addition to：Using the method for matrix decomposition, completion is carried out to the incomplete value in the consumption rating matrix.
- 5. method as claimed in claim 4, it is characterised in that the method using matrix decomposition, score the consumption Incomplete value in matrix carries out completion, including：First parameter line vector sum the second parameter row vector of random generation, the element number of the first parameter row vector with it is described The line number of consumption rating matrix is equal, the columns phase of the element number of the second parameter row vector and the consumption rating matrix Deng；The second parameter row vector according to the first parameter line vector sum, calculate the error of the consumption rating matrix；Second parameter row vector described in the first parameter line vector sum according to the error update, and repeat step is according to Second parameter row vector described in first parameter line vector sum, the error of the consumption rating matrix is calculated, until the error is received Hold back；The second parameter row vector determines the consumption rating matrix after completion according to the first parameter line vector sum.
- 6. the method as described in claim 1, it is characterised in that it is described word corresponding to the behavioral data is mapped to it is described Consume on a transaction dimension described in rating matrix with reference to user, including：For a transaction dimension with reference to user, transaction dimension word corresponding with the behavioral data is calculated Similarity；From the All Activity dimension with reference to user, it is determined that the similarity highest of word corresponding with the behavioral data Transaction dimension；Word corresponding to the behavioral data is mapped on the similarity highest transaction dimension.
- 7. the method as described in claim 1, it is characterised in that the quantity with reference to user is multiple；It is described that the financial label with reference to user is determined according to the comprehensive grading with reference to user, including：According to business rule and the background information with reference to user, from all with reference to the reference for determining to belong to same class label in user User；According to the comprehensive grading for the reference user for belonging to same class label, the forecast model of such label is determined；According to the forecast model of all kinds of labels, obtain integrating labeling model；According to the comprehensive grading with reference to user and the comprehensive labeling model, the finance mark with reference to user is determined Label.
- 8. the method as described in any one of claim 1 to 7, it is characterised in that described according to the behavioral data and mapping The consumption scoring of transaction dimension, after determining the comprehensive grading with reference to user, in addition to：It is determined that forming the historical time of comprehensive grading matrix, the comprehensive grading matrix is all comprehensive grading groups with reference to user Into；According to the historical time and current time, calculate under the current time, the comprehensive grading matrix after decay；Comprehensive grading matrix after the decay is calculated according to below equation,<mrow> <msup> <mi>M</mi> <mo>,</mo> </msup> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfrac> <mi>&alpha;</mi> <mrow> <mi>&alpha;</mi> <mo>+</mo> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mi>T</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mi>M</mi> <mrow> <mo>(</mo> <mi>T</mi> <mo>)</mo> </mrow> <mo>+</mo> <mi>M</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mrow>Wherein, α is decay factor, and t is current time, and T is historical time, and M (T) is the comprehensive grading matrix under historical time, M (t) it is the comprehensive grading matrix under current time, M ' (t) is the comprehensive grading matrix after the decay.
- A kind of 9. construction device of financial label, it is characterised in that including：Acquiring unit, for obtaining transaction data and user behaviors log with reference to user；Transaction handling unit, for according to the transaction data with reference to user, structure consumption rating matrix, the consumption scoring An element in matrix is the consumption scoring with reference to user on a transaction dimension；Text-processing unit, for according to the user behaviors log with reference to user, establishing the vector space mould with reference to user Type, the vector space model include multiple behavioral datas with reference to user, and each behavioral data corresponds to the reference A word in the user behaviors log of user；Computing unit is combined, for for a behavioral data with reference to user, by word corresponding to the behavioral data It is mapped on a transaction dimension described in the consumption rating matrix with reference to user, according to the behavioral data and mapping The consumption scoring of transaction dimension, determines the comprehensive grading with reference to user；Tag unit, for according to the comprehensive grading with reference to user, determining the financial label with reference to user.
- 10. device as claimed in claim 9, it is characterised in that the transaction handling unit, be specifically used for：User is referred to for each, using the transaction data with reference to user, calculates the reference user in different transaction dimensions The condition of consumption of degree；According to the condition of consumption, calculating is described to score with reference to user in the consumption of each transaction dimension；Scored using all with reference to user in the consumption of each transaction dimension, structure consumption rating matrix.
- 11. device as claimed in claim 10, it is characterised in that the transaction handling unit, specifically for utilizing following public affairs Formula calculating is described to score with reference to user in the consumption of a transaction dimension：Wherein, Score is consumption scoring of the user in a transaction dimension, and θ is the weight of the transaction dimension；ω is institute State with reference to user in the consumption stroke count of the transaction dimension and the weighted average of spending amount；υ is all reference users in institute State the consumption average of transaction dimension, σ be it is all with reference to user the transaction dimension variance；Be it is described with reference to user in institute State the spending amount of transaction dimension and the ratio of all spending amount sums with reference to user.
- 12. device as claimed in claim 9, it is characterised in that the transaction handling unit, be additionally operable to：Using the method for matrix decomposition, completion is carried out to the incomplete value in the consumption rating matrix.
- 13. device as claimed in claim 12, it is characterised in that the transaction handling unit, be specifically used for：First parameter line vector sum the second parameter row vector of random generation, the element number of the first parameter row vector with it is described The line number of consumption rating matrix is equal, the columns phase of the element number of the second parameter row vector and the consumption rating matrix Deng；The second parameter row vector according to the first parameter line vector sum, calculate the error of the consumption rating matrix；Second parameter row vector described in the first parameter line vector sum according to the error update, and repeat step is according to Second parameter row vector described in first parameter line vector sum, the error of the consumption rating matrix is calculated, until the error is received Hold back；The second parameter row vector determines the consumption rating matrix after completion according to the first parameter line vector sum.
- 14. device as claimed in claim 9, it is characterised in that the combination computing unit, be specifically used for：For a transaction dimension with reference to user, transaction dimension word corresponding with the behavioral data is calculated Similarity；From the All Activity dimension with reference to user, it is determined that the similarity highest of word corresponding with the behavioral data Transaction dimension；Word corresponding to the behavioral data is mapped on the similarity highest transaction dimension.
- 15. device as claimed in claim 9, it is characterised in that the quantity with reference to user is multiple；The tag unit, is specifically used for：According to business rule and the background information with reference to user, from all with reference to the reference for determining to belong to same class label in user User；According to the comprehensive grading for the reference user for belonging to same class label, the forecast model of such label is determined；According to the forecast model of all kinds of labels, obtain integrating labeling model；According to the comprehensive grading with reference to user and the comprehensive labeling model, the finance mark with reference to user is determined Label.
- 16. the device as described in any one of claim 9 to 15, it is characterised in that the combination computing unit, be additionally operable to：It is determined that forming the historical time of comprehensive grading matrix, the comprehensive grading matrix is all comprehensive grading groups with reference to user Into；According to the historical time and current time, calculate under the current time, the comprehensive grading matrix after decay；Comprehensive grading matrix after the decay is calculated according to below equation,<mrow> <msup> <mi>M</mi> <mo>,</mo> </msup> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfrac> <mi>&alpha;</mi> <mrow> <mi>&alpha;</mi> <mo>+</mo> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mi>T</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mi>M</mi> <mrow> <mo>(</mo> <mi>T</mi> <mo>)</mo> </mrow> <mo>+</mo> <mi>M</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mrow>Wherein, α is decay factor, and t is current time, and T is historical time, and M (T) is the comprehensive grading matrix under historical time, M (t) it is the comprehensive grading matrix under current time, M ' (t) is the comprehensive grading matrix after the decay.
- A kind of 17. computing device, it is characterised in that including：Memory, instructed for storage program；Processor, for calling the programmed instruction stored in the memory, performed according to the program of acquisition：Acquisition refers to user Transaction data and user behaviors log；According to the transaction data with reference to user, structure consumption rating matrix, the consumption is commented An element in sub-matrix is the consumption scoring with reference to user on a transaction dimension；According to described with reference to user's User behaviors log, establishes the vector space model with reference to user, and the vector space model includes described with reference to user's Multiple behavioral datas, each behavioral data correspond to a word in the user behaviors log with reference to user；For the reference The behavioral data of user, word corresponding to the behavioral data is mapped to described in the consumption rating matrix with reference to use On one transaction dimension at family, scored, determined described with reference to use according to the consumption of the behavioral data and the transaction dimension of mapping The comprehensive grading at family；According to the comprehensive grading with reference to user, the financial label with reference to user is determined.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710655552.XA CN107578270A (en) | 2017-08-03 | 2017-08-03 | A kind of construction method, device and the computing device of financial label |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710655552.XA CN107578270A (en) | 2017-08-03 | 2017-08-03 | A kind of construction method, device and the computing device of financial label |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107578270A true CN107578270A (en) | 2018-01-12 |
Family
ID=61035073
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710655552.XA Pending CN107578270A (en) | 2017-08-03 | 2017-08-03 | A kind of construction method, device and the computing device of financial label |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107578270A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109035028A (en) * | 2018-06-29 | 2018-12-18 | 平安科技（深圳）有限公司 | Intelligence, which is thrown, cares for strategy-generating method and device, electronic equipment, storage medium |
| CN109034941A (en) * | 2018-06-13 | 2018-12-18 | 平安科技（深圳）有限公司 | Products Show method, apparatus, computer equipment and storage medium |
| CN109543668A (en) * | 2018-11-29 | 2019-03-29 | 税友软件集团股份有限公司 | A kind of salary bill item identification method, device, equipment and readable storage medium storing program for executing |
| CN109614606A (en) * | 2018-10-23 | 2019-04-12 | 中山大学 | Long article this case fine range classification prediction technique and device based on document insertion |
| CN109635990A (en) * | 2018-10-12 | 2019-04-16 | 阿里巴巴集团控股有限公司 | A kind of training method, prediction technique, device and electronic equipment |
| CN110019563A (en) * | 2018-08-09 | 2019-07-16 | 北京首钢自动化信息技术有限公司 | A kind of portrait modeling method and device based on multidimensional data |
| CN110210692A (en) * | 2018-05-24 | 2019-09-06 | 腾讯科技（深圳）有限公司 | A kind of trade company's mode identification method, device, storage medium and terminal device |
| WO2019223379A1 (en) * | 2018-05-22 | 2019-11-28 | 阿里巴巴集团控股有限公司 | Product recommendation method and device |
| WO2019232891A1 (en) * | 2018-06-06 | 2019-12-12 | 平安科技（深圳）有限公司 | Method and device for acquiring user portrait, computer apparatus and storage medium |
| CN111414122A (en) * | 2019-12-26 | 2020-07-14 | 腾讯科技（深圳）有限公司 | Intelligent text processing method and device, electronic equipment and storage medium |
| CN112766293A (en) * | 2019-11-05 | 2021-05-07 | 腾讯科技（深圳）有限公司 | Data feature extraction method and business object classification method and device |
| CN113506138A (en) * | 2021-07-16 | 2021-10-15 | 瑞幸咖啡信息技术(厦门)有限公司 | Data estimation method, device, equipment and storage medium of business object |
| CN113963367A (en) * | 2021-10-22 | 2022-01-21 | 深圳前海环融联易信息科技服务有限公司 | Financial transaction file based on model and money extraction method |
| TWI779387B (en) * | 2020-11-06 | 2022-10-01 | 台北富邦商業銀行股份有限公司 | Smart customer tagging device and method thereof |
| CN117708183A (en) * | 2023-11-08 | 2024-03-15 | 广州西米科技有限公司 | Potential user mining method and system based on user consumption habit |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130073336A1 (en) * | 2011-09-15 | 2013-03-21 | Stephan HEATH | System and method for using global location information, 2d and 3d mapping, social media, and user behavior and information for a consumer feedback social media analytics platform for providing analytic measfurements data of online consumer feedback for global brand products or services of past, present, or future customers, users or target markets |
| CN103295145A (en) * | 2012-02-28 | 2013-09-11 | 北京星源无限传媒科技有限公司 | Mobile phone advertising method based on user consumption feature vector |
| CN105959745A (en) * | 2016-05-25 | 2016-09-21 | 北京铭嘉实咨询有限公司 | Advertising method and system |
| CN106022869A (en) * | 2016-05-12 | 2016-10-12 | 北京邮电大学 | Consumption object recommending method and consumption object recommending device |
-
2017
- 2017-08-03 CN CN201710655552.XA patent/CN107578270A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130073336A1 (en) * | 2011-09-15 | 2013-03-21 | Stephan HEATH | System and method for using global location information, 2d and 3d mapping, social media, and user behavior and information for a consumer feedback social media analytics platform for providing analytic measfurements data of online consumer feedback for global brand products or services of past, present, or future customers, users or target markets |
| CN103295145A (en) * | 2012-02-28 | 2013-09-11 | 北京星源无限传媒科技有限公司 | Mobile phone advertising method based on user consumption feature vector |
| CN106022869A (en) * | 2016-05-12 | 2016-10-12 | 北京邮电大学 | Consumption object recommending method and consumption object recommending device |
| CN105959745A (en) * | 2016-05-25 | 2016-09-21 | 北京铭嘉实咨询有限公司 | Advertising method and system |
Non-Patent Citations (1)
| Title |
|---|
| 高艳等: "面向用户偏好发现的隐变量模型构建与推理", 《计算机应用》 * |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019223379A1 (en) * | 2018-05-22 | 2019-11-28 | 阿里巴巴集团控股有限公司 | Product recommendation method and device |
| CN110210692A (en) * | 2018-05-24 | 2019-09-06 | 腾讯科技（深圳）有限公司 | A kind of trade company's mode identification method, device, storage medium and terminal device |
| WO2019232891A1 (en) * | 2018-06-06 | 2019-12-12 | 平安科技（深圳）有限公司 | Method and device for acquiring user portrait, computer apparatus and storage medium |
| CN109034941B (en) * | 2018-06-13 | 2023-03-31 | 平安科技（深圳）有限公司 | Product recommendation method and device, computer equipment and storage medium |
| CN109034941A (en) * | 2018-06-13 | 2018-12-18 | 平安科技（深圳）有限公司 | Products Show method, apparatus, computer equipment and storage medium |
| CN109035028B (en) * | 2018-06-29 | 2023-08-22 | 平安科技（深圳）有限公司 | Intelligent consultation strategy generation method and device, electronic equipment and storage medium |
| WO2020000689A1 (en) * | 2018-06-29 | 2020-01-02 | 平安科技（深圳）有限公司 | Transfer-learning-based robo-advisor strategy generation method and apparatus, and electronic device and storage medium |
| CN109035028A (en) * | 2018-06-29 | 2018-12-18 | 平安科技（深圳）有限公司 | Intelligence, which is thrown, cares for strategy-generating method and device, electronic equipment, storage medium |
| CN110019563A (en) * | 2018-08-09 | 2019-07-16 | 北京首钢自动化信息技术有限公司 | A kind of portrait modeling method and device based on multidimensional data |
| CN109635990B (en) * | 2018-10-12 | 2022-09-16 | 创新先进技术有限公司 | Training method, prediction method, device, electronic equipment and storage medium |
| CN109635990A (en) * | 2018-10-12 | 2019-04-16 | 阿里巴巴集团控股有限公司 | A kind of training method, prediction technique, device and electronic equipment |
| CN109614606B (en) * | 2018-10-23 | 2023-02-03 | 中山大学 | Document embedding-based long text case penalty range classification prediction method and device |
| CN109614606A (en) * | 2018-10-23 | 2019-04-12 | 中山大学 | Long article this case fine range classification prediction technique and device based on document insertion |
| CN109543668A (en) * | 2018-11-29 | 2019-03-29 | 税友软件集团股份有限公司 | A kind of salary bill item identification method, device, equipment and readable storage medium storing program for executing |
| CN112766293A (en) * | 2019-11-05 | 2021-05-07 | 腾讯科技（深圳）有限公司 | Data feature extraction method and business object classification method and device |
| CN112766293B (en) * | 2019-11-05 | 2024-08-09 | 腾讯科技（深圳）有限公司 | Data feature extraction method, business object classification method and device |
| CN111414122A (en) * | 2019-12-26 | 2020-07-14 | 腾讯科技（深圳）有限公司 | Intelligent text processing method and device, electronic equipment and storage medium |
| CN111414122B (en) * | 2019-12-26 | 2021-06-11 | 腾讯科技（深圳）有限公司 | Intelligent text processing method and device, electronic equipment and storage medium |
| TWI779387B (en) * | 2020-11-06 | 2022-10-01 | 台北富邦商業銀行股份有限公司 | Smart customer tagging device and method thereof |
| CN113506138B (en) * | 2021-07-16 | 2024-06-07 | 瑞幸咖啡信息技术(厦门)有限公司 | Data prediction method, device and equipment of business object and storage medium |
| CN113506138A (en) * | 2021-07-16 | 2021-10-15 | 瑞幸咖啡信息技术(厦门)有限公司 | Data estimation method, device, equipment and storage medium of business object |
| CN113963367B (en) * | 2021-10-22 | 2024-05-28 | 深圳前海环融联易信息科技服务有限公司 | Model-based financial transaction file and money extraction method |
| CN113963367A (en) * | 2021-10-22 | 2022-01-21 | 深圳前海环融联易信息科技服务有限公司 | Financial transaction file based on model and money extraction method |
| CN117708183A (en) * | 2023-11-08 | 2024-03-15 | 广州西米科技有限公司 | Potential user mining method and system based on user consumption habit |
| CN117708183B (en) * | 2023-11-08 | 2024-06-11 | 广州西米科技有限公司 | Potential user mining method and system based on user consumption habit |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107578270A (en) | A kind of construction method, device and the computing device of financial label | |
| Backhaus et al. | Multivariate analysis | |
| Jones et al. | An empirical evaluation of the performance of binary classifiers in the prediction of credit ratings changes | |
| Wang et al. | Mining semantic soft factors for credit risk evaluation in peer-to-peer lending | |
| Brezigar-Masten et al. | CART-based selection of bankruptcy predictors for the logit model | |
| Narayan et al. | The relationship between the real exchange rate and balance of payments: empirical evidence for China from cointegration and causality testing | |
| Piñeiro-Chousa et al. | Examining the influence of stock market variables on microblogging sentiment | |
| Bravo et al. | Granting and managing loans for micro-entrepreneurs: New developments and practical experiences | |
| CN110599336B (en) | Financial product purchase prediction method and system | |
| CN107480187A (en) | User's value category method and apparatus based on cluster analysis | |
| CN104778186B (en) | Merchandise items are mounted to the method and system of standardized product unit | |
| CN106611375A (en) | Text analysis-based credit risk assessment method and apparatus | |
| US20200234218A1 (en) | Systems and methods for entity performance and risk scoring | |
| Kibekbaev et al. | Benchmarking regression algorithms for income prediction modeling | |
| CN106294882A (en) | Data digging method and device | |
| Thewissen et al. | Unpacking the black box of ICO white papers: A topic modeling approach | |
| CN109740642A (en) | Invoice category recognition methods, device, electronic equipment and readable storage medium storing program for executing | |
| Caylor et al. | Analysts' qualitative statements and the profitability of favorable investment recommendations | |
| Chhabra | Does service quality matters in the context of internet banking: a perceptual analysis of Indias internet banking customers perception regarding service quality, trust, satisfaction and loyalty | |
| CN111754287A (en) | Article screening method, apparatus, device and storage medium | |
| Shin et al. | Determining directions of service quality management using online review mining with interpretable machine learning | |
| Soltani et al. | Ranking decision making units based on the multi-directional efficiency measure | |
| Tiwari et al. | Market segmentation using supervised and unsupervised learning techniques for E-commerce applications | |
| Dong et al. | Time to delisted status for listed firms in Chinese stock markets: An analysis using a mixture cure model with time-varying covariates | |
| Malyscheff et al. | Natural gas storage valuation via least squares Monte Carlo and support vector regression |
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 |
Application publication date: 20180112 |
|
| RJ01 | Rejection of invention patent application after publication |