CN104598780A

CN104598780A - Account identification method and system

Info

Publication number: CN104598780A
Application number: CN201310533508.3A
Authority: CN
Inventors: 杜玮; 朱训
Original assignee: Alibaba Group Holding Ltd
Current assignee: Advanced New Technologies Co Ltd
Priority date: 2013-10-31
Filing date: 2013-10-31
Publication date: 2015-05-06
Anticipated expiration: 2033-10-31
Also published as: HK1205580A1; CN104598780B

Abstract

The invention discloses an account identification method and system. The account identification method comprises the steps of obtaining information uniformized chain related information of the same attributes of accounts to be identified; conducting first identification judgment on the accounts to be identified according to the information uniformized chain related information; judging whether the accounts to be identified are used by users corresponding to the accounts to be identified according to results of the information uniformized chain related information at least, wherein the information uniformized chain related information include one of the length of an information uniformized chains, the number of predetermined important indexes included in the information uniformized chains or the length information uniformized chains, the number of the predetermined important indexes included in the information uniformized chain and coincidence indexes corresponding to the information uniformized chains. By means of a target-free modeling method, the problem that an existing account identification method adopts a destination training model and accordingly needs to depend on a large amount of fact sample data is solved.

Description

Account recognition methods and system

Technical field

The application relates to internet security field, in particular to a kind of account recognition methods and system.

Background technology

Usually adopt the method for targetedly training pattern to realize the identification of the user to account in prior art, typical method comprises the most:

1) decision tree---first data are processed, utilize inductive algorithm to generate readable rule and decision tree, then use decision-making to analyze new data.Decision tree is the process of being classified to data by series of rules in essence.

2) scorecard model---first according to the historical summary of user, utilize certain credit scoring model or algorithm, obtain mark or the sequence of different brackets, then decide customer risk level according to the mark of user or sequence.

Above-mentioned two kinds targetedly the method for training pattern usually need to gather true sample data in advance, also be, need to know account whether the historical facts record that uses by the user corresponding with the account, such as, can by the collection of complaint realization to true sample data of customer phone, then rely on the true sample data collected to provide input and output for model simultaneously, realize the training to model.But, because the feature gathering the sample come like this is usually not representative, thus cause the accuracy of the model of cognition drawn usually poor.In addition; modeling period in this account recognition methods is usually longer, and when the use habit of user group changes in time, usually needs modeling again; be modeling or modeling all causes extra expense to system again in a word, cause the waste of system resource.

For the problems referred to above, at present effective solution is not yet proposed.

Summary of the invention

The fundamental purpose of the application is to provide a kind of account recognition methods and system, at least to solve the problem that the needs adopting the mode of targetedly training pattern to cause due to existing account recognition methods rely on a large amount of true sample datas.

According to an aspect of the application, provide a kind of account recognition methods, comprising: the consistent chain relevant information of information obtaining the same attribute of account to be identified, and according to above-mentioned information consistent chain relevant information, the first identification judgement is carried out to above-mentioned account to be identified; At least identify that the result judged judges whether above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified according to above-mentioned first; Wherein, above-mentioned information consistent chain relevant information comprises one of following: the length of the consistent chain of information, wherein, the consistent chain of above-mentioned information comprises and belongs to above-mentioned same attribute and the log-on message of the content user corresponding with above-mentioned account to be identified one or more indexs consistent with each other, and the length of the consistent chain of above-mentioned information represents the number of the These parameters that the consistent chain of above-mentioned information comprises; The number of predetermined important indicator is comprised in the length chain consistent with above-mentioned information of the consistent chain of above-mentioned information; Or comprise the same index that the number chain consistent with above-mentioned information of predetermined important indicator is corresponding in the length of the consistent chain of above-mentioned information, the consistent chain of above-mentioned information.

Alternatively, above-mentioned attribute comprises multiple, wherein, above-mentionedly according to above-mentioned information consistent chain relevant information, the first identification is carried out to above-mentioned account to be identified and judge to comprise: obtain attribute ratings corresponding to each attribute of each above-mentioned account to be identified according to above-mentioned information consistent chain relevant information, wherein, the attribute ratings that above-mentioned attribute is corresponding represents the relative probability that the account above-mentioned to be identified that goes out based on above-mentioned determined property is used by the user corresponding with above-mentioned account to be identified; The range identification grade of each above-mentioned account to be identified is judged according to the distribution situation of all above-mentioned accounts to be identified on each probability levels of each above-mentioned attribute, wherein, the range identification grade of above-mentioned account to be identified represents the relative probability that above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified.

Alternatively, above-mentioned each attribute comprises the first attribute, wherein, the above-mentioned attribute ratings corresponding according to each attribute of each above-mentioned account to be identified of above-mentioned information consistent chain relevant information acquisition comprises: the attribute ratings corresponding according to above-mentioned first attribute of each above-mentioned account to be identified of the distributed acquisition of the length of the consistent chain of information of the first attribute of all above-mentioned accounts to be identified, wherein, for the first account to be identified that the length of the consistent chain of above-mentioned information in all above-mentioned accounts to be identified is identical, the attribute ratings corresponding according to above-mentioned first attribute of above-mentioned first account to be identified of number acquisition comprising predetermined important indicator in the distribution of the length of the consistent chain of information of the first attribute of above-mentioned first account to be identified and the consistent chain of information of above-mentioned first attribute of above-mentioned first account to be identified, and second to be identified account that the above-mentioned number that comprise predetermined important indicator identical identical for the length of the consistent chain of above-mentioned information in all above-mentioned accounts to be identified, the same index corresponding according to the number and the consistent chain of above-mentioned information that comprise predetermined important indicator in the distribution of the length of the consistent chain of information of the first attribute of above-mentioned second account to be identified and the consistent chain of information of above-mentioned first attribute of above-mentioned second account to be identified obtains attribute ratings corresponding to above-mentioned first attribute of above-mentioned second account to be identified.

Alternatively, above-mentioned at least identify that the result judged judges that whether above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified according to above-mentioned first before, above-mentioned account recognition methods comprises: obtain the accounting that the active time of above-mentioned account to be identified and above-mentioned account to be identified are used by the user not corresponding with above-mentioned account to be identified, and carry out the second identification judgement according to above-mentioned active time and above-mentioned accounting to above-mentioned account to be identified; Wherein, above-mentionedly at least identify that the result judged judges that whether above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified and comprises according to above-mentioned first: identify that the result and above-mentioned second judged identifies that the result judged judges whether above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified according to above-mentioned first.

Alternatively, above-mentionedly according to above-mentioned active time and above-mentioned accounting, the second identification is carried out to above-mentioned account to be identified and judge to comprise: what obtain each above-mentioned account to be identified according to the active time of each above-mentioned account to be identified enlivens grade, and is obtained the accounting grade of each above-mentioned account to be identified by the accounting that the user not corresponding with above-mentioned account to be identified uses according to each above-mentioned account to be identified; Judge the depth recognition grade of each above-mentioned account to be identified in each above-mentioned distribution situation enlivened in grade and above-mentioned each accounting grade according to all above-mentioned accounts to be identified, wherein, the depth recognition grade of above-mentioned account to be identified represents the relative probability that above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified.

Alternatively, the grade of enlivening that the above-mentioned active time according to each above-mentioned account to be identified obtains each above-mentioned account to be identified comprises: what obtain each above-mentioned account to be identified according to the ratio between active time the longest in the active time of each above-mentioned account to be identified and all above-mentioned accounts to be identified enlivens grade, wherein, the difference belonged between each above-mentioned number enlivening the account above-mentioned to be identified of grade is less than predetermined threshold.

Alternatively, above-mentionedly to be comprised by the accounting grade that the accounting that the user not corresponding with above-mentioned account to be identified uses obtains each above-mentioned account to be identified according to each above-mentioned account to be identified: the accounting grade obtaining each above-mentioned account to be identified according to the ratio between accounting maximum in the above-mentioned accounting of each above-mentioned account to be identified and all above-mentioned accounts to be identified, wherein, the difference between the number belonging to the account above-mentioned to be identified of each above-mentioned accounting grade is less than predetermined threshold.

Alternatively, above-mentionedly identify that the result and above-mentioned second judged identifies that the result judged judges that whether above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified and comprises according to above-mentioned first: obtain according to the range identification grade of each above-mentioned account to be identified and depth recognition grade the relative probability that the user corresponding with above-mentioned account to be identified use.

According to the another aspect of the application, additionally provide a kind of account recognition system, comprise: the first recognition unit, the consistent chain relevant information of the information for the same attribute obtaining account to be identified, and according to above-mentioned information consistent chain relevant information, the first identification judgement is carried out to above-mentioned account to be identified; According to above-mentioned first, judging unit, at least identifying that the result judged judges whether above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified; Wherein, above-mentioned information consistent chain relevant information comprises one of following: the length of the consistent chain of information, wherein, the consistent chain of above-mentioned information comprises and belongs to above-mentioned same attribute and the log-on message of the content user corresponding with above-mentioned account to be identified one or more indexs consistent with each other, and the length of the consistent chain of above-mentioned information represents the number of the These parameters that the consistent chain of above-mentioned information comprises; The number of predetermined important indicator is comprised in the length chain consistent with above-mentioned information of the consistent chain of above-mentioned information; Or comprise the same index that the number chain consistent with above-mentioned information of predetermined important indicator is corresponding in the length of the consistent chain of above-mentioned information, the consistent chain of above-mentioned information.

Alternatively, above-mentioned attribute comprises multiple, wherein, above-mentioned first recognition unit comprises: the first acquisition module, for the attribute ratings that each attribute obtaining each above-mentioned account to be identified according to above-mentioned information consistent chain relevant information is corresponding, wherein, the attribute ratings that above-mentioned attribute is corresponding represents the relative probability that the account above-mentioned to be identified that goes out based on above-mentioned determined property is used by the user corresponding with above-mentioned account to be identified; First judge module, for judging the range identification grade of each above-mentioned account to be identified according to the distribution situation of all above-mentioned accounts to be identified on each probability levels of each above-mentioned attribute, wherein, the range identification grade of above-mentioned account to be identified represents the relative probability that above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified.

Alternatively, above-mentioned each attribute comprises the first attribute, wherein, above-mentioned first acquisition module comprises: first obtains submodule, the attribute ratings that above-mentioned first attribute of each above-mentioned account to be identified of the distributed acquisition for the length of the consistent chain of information of the first attribute according to all above-mentioned accounts to be identified is corresponding, wherein, for the first account to be identified that the length of the consistent chain of above-mentioned information in all above-mentioned accounts to be identified is identical, the attribute ratings corresponding according to above-mentioned first attribute of above-mentioned first account to be identified of number acquisition comprising predetermined important indicator in the distribution of the length of the consistent chain of information of the first attribute of above-mentioned first account to be identified and the consistent chain of information of above-mentioned first attribute of above-mentioned first account to be identified, and second to be identified account that the above-mentioned number that comprise predetermined important indicator identical identical for the length of the consistent chain of above-mentioned information in all above-mentioned accounts to be identified, the same index corresponding according to the number and the consistent chain of above-mentioned information that comprise predetermined important indicator in the distribution of the length of the consistent chain of information of the first attribute of above-mentioned second account to be identified and the consistent chain of information of above-mentioned first attribute of above-mentioned second account to be identified obtains attribute ratings corresponding to above-mentioned first attribute of above-mentioned second account to be identified.

Alternatively, above-mentioned account recognition system also comprises: the second recognition unit, for the accounting that the active time and above-mentioned account to be identified that obtain above-mentioned account to be identified are used by the user not corresponding with above-mentioned account to be identified, and according to above-mentioned active time and above-mentioned accounting, the second identification judgement is carried out to above-mentioned account to be identified; Wherein, above-mentioned judging unit is used for identifying that the result and above-mentioned second judged identifies that the result judged judges whether above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified according to above-mentioned first.

Alternatively, above-mentioned second recognition unit comprises: the second acquisition module, enliven grade for what obtain each above-mentioned account to be identified according to the active time of each above-mentioned account to be identified, and obtained the accounting grade of each above-mentioned account to be identified by the accounting that the user not corresponding with above-mentioned account to be identified uses according to each above-mentioned account to be identified; Second judge module, for judging the depth recognition grade of each above-mentioned account to be identified in each above-mentioned distribution situation enlivened in grade and above-mentioned each accounting grade according to all above-mentioned accounts to be identified, wherein, the depth recognition grade of above-mentioned account to be identified represents the relative probability that above-mentioned account to be identified is used by the user corresponding with above-mentioned account to be identified.

Alternatively, above-mentioned second acquisition module comprises: second obtains submodule, grade is enlivened for what obtain each above-mentioned account to be identified according to the ratio between active time the longest in the active time of each above-mentioned account to be identified and all above-mentioned accounts to be identified, wherein, the difference belonged between each above-mentioned number enlivening the account above-mentioned to be identified of grade is less than predetermined threshold.

Alternatively, above-mentioned second acquisition module comprises: the 3rd obtains submodule, for obtaining the accounting grade of each above-mentioned account to be identified according to the ratio between accounting maximum in the above-mentioned accounting of each above-mentioned account to be identified and all above-mentioned accounts to be identified, wherein, the difference between the number belonging to the account above-mentioned to be identified of each above-mentioned accounting grade is less than predetermined threshold.

Alternatively, above-mentioned judging unit comprises: the 3rd judge module, for the relative probability used according to range identification grade and the depth recognition grade acquisition user corresponding with above-mentioned account to be identified of each above-mentioned account to be identified.

In the embodiment of the present application, by the retrieval of relevant information record recorded account process different business to be identified, judge the range identification grade in the range of process business of account to be identified and the degree of depth of service time and depth recognition grade, and then realize the differentiation to the identity of the user of account to be identified by range identification grade and/or depth recognition grade.In this application, owing to have employed aimless modeling pattern, thus the dependence eliminated true sample data, and then solve the problem that the needs adopting the mode of targetedly training pattern to cause due to existing account recognition methods rely on a large amount of true sample datas.

In addition, in the embodiment of the present application, by the introducing to predetermined important indicator and same index, the correction to judgment models can also be realized further, thus reaches the technique effect of the accuracy promoting account identification further.Certainly, the arbitrary product implementing the application might not need to reach above-described all advantages simultaneously.

Accompanying drawing explanation

Accompanying drawing described herein is used to provide further understanding of the present application, and form a application's part, the schematic description and description of the application, for explaining the application, does not form the improper restriction to the application.In the accompanying drawings:

Fig. 1 is the schematic diagram of a kind of optional account recognition system according to the embodiment of the present application;

Fig. 2 is the schematic diagram according to the optional account recognition system of the another kind of the embodiment of the present application;

Fig. 3 is the schematic diagram of another the optional account recognition system according to the embodiment of the present application;

Fig. 4 is the schematic diagram of another the optional account recognition system according to the embodiment of the present application;

Fig. 5 is the schematic diagram of another the optional account recognition system according to the embodiment of the present application;

Fig. 6 is the schematic diagram of another the optional account recognition system according to the embodiment of the present application;

Fig. 7 is the schematic diagram of a kind of optional account recognition methods according to the embodiment of the present application;

Fig. 8 is the schematic diagram according to the optional account recognition methods of the another kind of the embodiment of the present application;

Fig. 9 is the schematic diagram of another the optional account recognition methods according to the embodiment of the present application;

Figure 10 is the schematic diagram of another the optional account recognition methods according to the embodiment of the present application;

Figure 11 is the schematic diagram of another the optional account recognition methods according to the embodiment of the present application;

Figure 12 is the schematic diagram of another the optional account recognition methods according to the embodiment of the present application;

Figure 13 is the schematic diagram of another the optional account recognition methods according to the embodiment of the present application;

Figure 14 is the schematic diagram of another the optional account recognition system according to the embodiment of the present application.

Embodiment

Hereinafter also describe the application in detail with reference to accompanying drawing in conjunction with the embodiments.It should be noted that, when not conflicting, the embodiment in the application and the feature in embodiment can combine mutually.

Embodiment 1

The embodiment of the present application provides a kind of account recognition system.As shown in Figure 1, this system can comprise:

1) the first recognition unit 102, the consistent chain relevant information of the information for the same attribute obtaining account to be identified, and according to information consistent chain relevant information, the first identification judgement is carried out to account to be identified;

2) according to first, judging unit 104, for identifying that the result judged judges whether account to be identified is used by the user corresponding with account to be identified;

First will be clear that, one of technical problems to be solved in this application, be to provide a kind of recognition system, by this recognition system, can to the account whether by the user corresponding with the account, the registered user of the such as account uses and judges, on the other hand, by the judged result that this system obtains, also can be used for reflecting the credibility of the account.Such as, this account recognition system can be applied to the scene of network payment, wherein, when server receives the payment request of a certain account, the account recognition system provided according to the embodiment of the present application can be utilized to judge the credibility of the account, whether to make the account further in the judgement used by user corresponding with it, and finally make by or refuse the decision-making of this payment request.For ease of describing, below the identification object of account recognition system is designated as account to be identified.

For the purpose of more effectively solving the problem, in the account recognition system that the embodiment of the present application provides, the concept of the consistent chain of the information that proposes.Particularly, for an account to be identified, the consistent chain of the information corresponding to it can comprise one or more index.In a consistent chain of information, one or more index should belong to above-mentioned same attribute, the content of these indexs should be consistent with each other, and the content of index should be corresponding with account to be identified the log-on message of user consistent, at least should comprise in the log-on message of the user corresponding with account to be identified the index belonging to this attribute in other words in these indexs.

Table 1

	Name
		Certification/log-on message	A
Credit card repayment information	A
		To receive information	A
Government utility is paid the fees	B

The example provided below in conjunction with table 1 is described in detail the consistent chain of information.Wherein, this account to be identified, for representing an account-related information record to be identified, for ease of describing, is designated as the first account by table 1 below.

Wherein, content in table 1 shown in first row: certification/log-on message, credit card repayment information, to receive information, government utility is paid the fees, relevant information record during first this collection of different business of account process can be represented, corresponding with it, content shown in table 1 secondary series: " A ", " A ", " A ", " B ", the name information that above-mentioned relevant information record comprises can be represented, such as, can be learnt by table 1, first account is when processing certification or registration business, such as, when the Account Registration of transaction platform or certification, the content of the name information recorded can be " A ", the content of the index of " name " this attribute is belonged to for " A " in the log-on message of also namely corresponding with account to be identified user, and the first account is at process government utility fee payment service, such as, during the Online Payment of telephone expenses, the content of the name information recorded can be " B ", wherein, although this index also belongs to " name " attribute, but its content " B " is inconsistent with the content " A " of above-mentioned and that account to be identified is corresponding user's registration information, therefore this index can't be registered as one of member of the consistent chain of information.

Wherein, the content " A " of above-mentioned name information or " B " can be inputted in client by the user when the above-mentioned business of the execution of the first account usually, but the application is not construed as limiting this, such as, in some embodiments of the application, the content of information also can call the database in server.

It should be noted that, the relevant information record recorded in the process of the above-mentioned business of the first account process, usually the collection of information items be combined to form by more than one item of information can be shown as, such as, for certification or registration business, the collection of information items corresponding with the first account can comprise the item of information of different attribute usually: name, the pet name, age, identification card number, address, telephone number, email address etc., and in other business, the concrete form of the single item of information in the collection of information items corresponding with the first account both can be similar text message, also other data type and structure can be shown, such as the business of receiving, its information of receiving also can comprise individual subscriber and arrange the information such as code.But, in the collection of information items recorded when use first account process different business, usually there is one or more item of information belonging to same attribute, name as escribed above, or other feasible information content, such as phone numbers etc.

In the embodiment of the present application, above-mentioned item of information can be considered as the index of user to be identified, wherein, namely an item of information can represent an index, and the index belonging to same attribute combines, namely can form message sense, such as in Table 1, the message sense that the index belonging to same attribute " name " is formed can be " AAAB ".Further, by targets match consistent for the content in message sense to together, namely the consistent chain of information can be formed, two the consistent chains of information existed in such as table 1 can be respectively: " AAA " and " B ", wherein, the log-on message of the user that the content " A " of first three index is corresponding with account to be identified is consistent with each other, and the log-on message of the content of last index " B " user corresponding with account to be identified is inconsistent, therefore in the embodiment of the present application, the consistent chain of the information counted is: " AAA ", and the length of the consistent chain of its information is 3.Generally speaking, the consistent index of foregoing can show as the index with identical parameter value usually, but the application is not construed as limiting this, and it can also define other conforming identification standards known to those skilled in the art.

Certainly, be more than a kind of example, can't form any restriction to the application, such as, in other embodiments of the application, included by These parameters also can not be item of information itself, but the database index record corresponding with item of information, etc.

As previously mentioned, the consistent chain of the information corresponding with the first account can exist different in size multiple, and wherein, the length of the consistent chain of information may be used for the number of index included in the consistent chain of expression information, and the length of the consistent chain of the shortest information can be 1, also namely comprise an index.On the other hand, for different attributes, the consistent chain of the information formed by corresponding index also can there are differences.But this and do not mean that restriction constituted to the application, the technical scheme of the application can be implemented based on the consistent chain of information that wherein attribute is formed arbitrarily.

It should be noted that in addition, the acquisition for the consistent chain of information and relevant information thereof described in the application, the operation such as record and the follow-up judgement made according to these information, both can perform at server, also can in client executing, can also perform at server and client side two ends, the application is not construed as limiting this.In addition, the index in above-mentioned relevant information record chain consistent with information also can be stored on client and/or server, and the application is not construed as limiting this.

Certainly, it is more than a kind of example, so that understand the concept of the consistent chain of information that the application proposes, and should not be construed as any unnecessary restriction is constituted to the application, such as, in some embodiments of the application, the situation shown in table 1 is different from, also can comprise the multinomial name information repeated in the relevant information record of one of the multinomial business of account process to be identified, also namely can comprise the multiple indexs belonging to same attribute.

In addition it should be noted that, although provide the optional building mode of one of the consistent chain of information in the above-described embodiments, but this building mode embodiment that to be not technical scheme necessary, such as, the selection of the index in the consistent chain of information is with reference to also can not different business handled by account to be identified, and such as, These parameters also can be chosen from the Visitor Logs of account to be identified under different domain name, or choose in the different classes of event that account to be identified triggers, etc.Particularly, the application is not limited in any way for the retrieval mode of concrete index and the building mode of the consistent chain of information.

On basis described above, in the embodiment of the present application, the consistent chain relevant information of information of the same attribute of account to be identified can be obtained in the first recognition unit 102, and according to information consistent chain relevant information, the first identification judgement is carried out to account to be identified.Wherein, information consistent chain relevant information can comprise one of following:

1) length of the consistent chain of information, wherein, the consistent chain of described information comprise belong to above-mentioned same attribute and one or more indexs that content is consistent with each other, the length of the consistent chain of information represents the number of the index that the consistent chain of information comprises;

2) number of predetermined important indicator is comprised in the length of the consistent chain of information chain consistent with information;

3) same index that the number chain consistent with information of predetermined important indicator is corresponding is comprised in the length of the consistent chain of information, the consistent chain of information.

As previously mentioned, in the embodiment of the present application, carry out that the first identification judges according to can be the consistent chain relevant information of information, and this information can comprise the length of the consistent chain of foregoing information, is also the number of the index in the consistent chain of information.Wherein, generally speaking, the length of the consistent chain of information is longer, then represent that the index that the consistent chain of this information comprises is more, the number of the index that also namely relevant to the account to be identified content belonging to a certain attribute is consistent is more, thus account to be identified can be considered as there is higher credibility, and can know by inference thus account to be identified the probability that uses by user corresponding with it higher, otherwise, if the length of the consistent chain of information is shorter, then can be considered as account to be identified and there is lower credibility, and then know by inference account to be identified by with it corresponding for the probability that uses lower.Generally speaking, in identifying judge first, usually the length of consistent chain can judge above-mentioned probability according to the longest information, but the application is not construed as limiting this.

Certainly, above-described embodiment provide and do not mean that in the embodiment of the present application first identify judge mode be unique, such as, in some embodiments of the application, first identifies that the foundation judged can also comprise the number of the predetermined important indicator comprised in the consistent chain of information.Wherein, predetermined important indicator may be used for the index of the outbalance represented in the consistent chain of information, and this index is whether important, usually can the source of attribute belonging to this index and/or index itself arrange, such as, certification or the log-on message of transaction platform can be come from, in the relevant information record of the outbalances such as credit card repayment information, the index belonging to the attribute of the outbalance such as name or identification card number is predetermined important indicator, and then, can identify in judgement first, for the predetermined important indicator in the consistent chain of information gives higher weight, also be, can the number of predetermined important indicator that comprises of the length of the consistent chain of combining information and the consistent chain of information judge, such as, in some cases, although the consistent chain length of information is shorter, but predetermined important indicator is wherein relatively many, then still can judge that the probability that account to be identified is used by user corresponding is with it higher accordingly.

More specifically, assignment can also be carried out according to the significance level of above-mentioned predetermined important indicator, and calculate same index corresponding to the consistent chain of information as reference, thus judged result that is more accurate, that be easy to post-treatment operations can be obtained according to the numerical value of same index.Certainly, this is a kind of example, and in the embodiment of the present application, same index also can have nothing to do with predetermined important indicator, but sets the standard on data drawn according to other factors, and the application is not limited in any way this.Such as, in some embodiments of the application, the all indexs belonging to same attribute in the relevant information record of this account to be identified can be retrieved, wherein, often there is once another index identical with the content of predetermined index, then do to add divisional processing, often there is once another index different from the content of predetermined index, then do deduction process, thus the gross score calculated can be identified as first the foundation judged as above-mentioned same index, wherein, the account to be identified that gross score is higher can be considered as by the probability of user corresponding with it use higher, otherwise it is lower then can be considered as this probability.

Known from the above description, the judged result judged is identified as first, the account to be identified that can draw for factors such as the length according to the consistent chain of information whether use by the user corresponding with account to be identified one relative probable value, such as, this probable value can directly be represented by above-mentioned same index, but, in some embodiments of the application, also the length according to the consistent chain of information can be shown as, the judged result that the number of predetermined important indicator and/or same index corresponding to the consistent chain of information draw via other processing procedures is comprised in the consistent chain of information, this judged result also can not be embodied in a certain probability numbers, the application is not limited in any way this.

In embodiments of the present invention, can also be further advanced by statistics more is accurately obtained to the consistent chain relevant information of the information of multiple attributes of user to be identified.To first, some embodiments by the application are identified that the optional decision logic judged makes a more detailed description below.

As shown in Figure 2, as a kind of optional embodiment of the application, the first recognition unit 102 can comprise:

1) the first acquisition module 202, for the attribute ratings that each attribute obtaining each described account to be identified according to described information consistent chain relevant information is corresponding, wherein, the attribute ratings that described attribute is corresponding represents the relative probability that the account described to be identified that goes out based on described determined property is used by the user corresponding with described account to be identified;

2) the first judge module 204, for judging the range identification grade of each described account to be identified according to the distribution situation of all described accounts to be identified on each probability levels of each described attribute, wherein, the range identification grade of described account to be identified represents the relative probability that described account to be identified is used by the user corresponding with described account to be identified.

In the embodiment of the present application, can after the consistent chain relevant information of above-mentioned information of a certain attribute obtaining multiple account to be identified, according to the distribution situation of the account to be identified in the different brackets interval of information consistent chain relevant information, for the credibility of the account to be identified on one of above-mentioned grade interval is demarcated, such as, the account to be identified that can be in the grade interval of a certain quantification for the consistent chain length of information is defined by the relative probability that user corresponding with it uses, and then can according to above-mentioned distribution situation or calibration result, the range identification grade judging account to be identified in judgement is identified first, wherein, namely this range identification grade may be used for the relative probability representing that this account to be identified is used by user corresponding with it.

Particularly, above-mentioned grade interval can the different information divided rank by the information being respectively a certain attribute in consistent chain relevant information realize, dividing mode such as shown in table 2, wherein, table 2 is with representing the attribute ratings marked off according to the length of the consistent chain of the information of the name information in relative recording information.

Table 2

The length L of the consistent chain of information	Attribute ratings L _A
		L<4	1
4≤L<10	2
		T≥10	3

As shown in table 2, in the embodiment of the present application, can to attribute ratings L _ado following division: the attribute ratings section definition being 1 to 3 by the length of consistent for information chain is attribute ratings L _aequaling 1, is that the attribute ratings section definition of 4 to 9 is L by length _aequaling 2, is that the attribute ratings section definition of more than 10 is L by length _aequal 3, thus the length that can realize the consistent chain of information of the name attribute according to each account to be identified in the first acquisition module 202 obtains the attribute ratings of each account to be identified, such as, if the length of the consistent chain of information of the name attribute of a certain account to be identified is 8, then can draw the attribute ratings L of this account to be identified _aequal 2.

Under above-mentioned scene, generally speaking, the account to be identified with higher attribute ratings can be considered as having higher credibility usually, thus can judge that the probability that account to be identified is used by user corresponding is with it higher, otherwise then can judge that probability is lower.But, in some embodiments of the application, be adopted as the deviation that mode that each attribute ratings gives the constant probable value of rigidity still may affect judged result and actual probabilities, thus the distribution situation of multiple accounts to be identified in each attribute ratings can be counted further in the first judge module 204, and then the depth recognition grade of each account to be identified can be judged according to this distribution situation.

Be described in detail below in conjunction with the deterministic process of table 3 to the range identification grade of account to be identified.Wherein, table 3 is for representing the quantity of the account to be identified that different attribute ratings distributes.

Table 3

The length L of the consistent chain of information	Attribute ratings L _A	Account quantity to be identified
			L<5	1	180
5≤L<10	2	200
			T≥10	3	30

In table 3, the distribution situation of above-mentioned account to be identified in each attribute ratings can show as the quantity of the account to be identified in each attribute ratings, such as, for L _a=1, also namely the statistics of the to be identified account of the consistent chain length of information within 4 can be 180.

On the basis of the above, and then can according to the distribution situation in table 3, judge by the first judge module 204 relative probability that account to be identified is used by user corresponding with it, also the range identification grade of i.e. this account to be identified.Such as, consider that the usage behavior of most of user is optimum, thus according to statistical law, it can be grade combination imparting forward adjustment factor that the distributed quantity of the user to be identified in table 3 is relatively high, and be the relatively low grade combination imparting negative regulation coefficient of the distributed quantity of user to be identified, wherein, forward and negative regulation coefficient can represent the probability used by user corresponding with it for improving and reduce the user to be identified that judges respectively, and then obtain user to be identified by the relative probability of user corresponding with it use according to above-mentioned adjustment factor, also be range identification grade.

Further, in some embodiments of the application, range identification grade can also judge in conjunction with the attribute ratings of multiple attribute, such as, shown in table 4, and wherein L _a-Nthe attribute ratings that " name " attribute of drawing according to the length of the consistent chain of information is corresponding can be represented, and L _a-Ithe attribute ratings that " identification card number " attribute of drawing according to the length of the consistent chain of information is corresponding can be represented.

Table 4

Under this scene, the division of attribute ratings can be floating, and such as, it also can correct according to the distribution of account to be identified in range identification grade.As shown in table 4, each L _a-Ngrade and L _a-Ithe quantity of the account to be identified corresponding to the combination of grade can be 50, also be evenly distributed in by account to be identified in the list formed by the combination of the attribute ratings of multiple attribute, can be all shorter situation of the consistent chain length of information that " name " and " identification card number " two attributes are corresponding, such as L _a-N=1 and L _a-Ithe account to be identified of=1 gives lower range identification grade, thus the lower account to be identified of correspondence the relative probability that uses by user corresponding with it, otherwise, can be L _a-N=3 and L _a-Ithe account to be identified of=3 gives higher range identification grade, correspond to higher account to be identified by the relative probability shown in user corresponding with it.

Further, in the embodiment of the present application, can also in conjunction with more attribute to construct the higher list of dimension to judge the range identification grade of account to be identified.On the other hand, outside the consistent chain length of information, can also construct the higher list of dimension in conjunction with according to information consistent chain relevant informations such as the number of the predetermined important attribute in the consistent chain of information or the same indexes of ratio and the consistent chain of information, the present invention is not limited in any way this.

Certainly, be more than some examples, in the embodiment of the present application, the decision logic of range identification grade can also have multiple, such as, can also respectively according to the length of the consistent chain of information, the number of predetermined important attribute in the consistent chain of information or ratio, and/or the same index of the consistent chain of information judges the range identification grade of account to be identified, and then, the range identification grade of account to be identified can also be judged according to the distribution situation in each combination of multiple attribute ratings of user to be identified between above-mentioned different information consistent chain relevant information, the application does not do tired stating at this.

And then as shown in Figure 3, in the embodiment of the present application, the first acquisition module 202 can also comprise:

1) first submodule 302 is obtained, the attribute ratings that the first attribute for each account to be identified of distributed acquisition of the length of the consistent chain of information of the first attribute according to all accounts to be identified is corresponding, wherein,

For the first account to be identified that the length of the consistent chain of information in all accounts to be identified is identical, obtain attribute ratings corresponding to the first attribute of the first account to be identified according to the number comprising predetermined important indicator in the distribution of the length of the consistent chain of information of the first attribute of the first account to be identified and the consistent chain of information of the first attribute of the first account to be identified;

Identical and comprise the second identical account to be identified of the number of predetermined important indicator for the length of the consistent chain of information in all accounts to be identified, the same index corresponding according to the number and the consistent chain of information that comprise predetermined important indicator in the distribution of the length of the consistent chain of information of the first attribute of the second account to be identified and the consistent chain of information of the first attribute of the second account to be identified obtains attribute ratings corresponding to the first attribute of the second account to be identified.

Wherein, for ease of statement, a certain attribute is designated as the first attribute.

On the other hand, in the embodiment of the present application, carry out the first identification judgement except the first recognition unit 102 can be utilized and in judging unit 104 according to first identify the result judged judge account to be identified whether use by the user corresponding with this account to be identified except, alternatively, as shown in Figure 4, above-mentioned account recognition system can also comprise:

1) the second recognition unit 402, the accounting that active time and account to be identified for obtaining account to be identified are used by the user not corresponding with account to be identified, and according to active time and accounting, the second identification judgement is carried out to account to be identified; Wherein,

According to first, judging unit 104 is for identifying that the result and second judged identifies that the result judged judges whether account to be identified is used by the user corresponding with account to be identified.

Wherein, be different from the first identification carried out according to the consistent chain relevant information of information to judge, second identifies that the basis for estimation that judges can as the active time of account to be identified, and account to be identified the accounting that uses by user not corresponding with it, non-ly in other words use accounting in person, wherein, this accounting may be used for representing the ratio of the overall access times of the number of times that the nonregistered user of account to be identified uses and this account to be identified usually, wherein, whether the name that the situation whether being undertaken using by registered user can be recorded by the user of account to be identified is consistent with each other with the name in the log-on message of account to be identified, but the application is not construed as limiting this, its can also for other feasible for represent account to be identified the quantized values of service condition that uses by user not corresponding with it.Correspondingly, in other embodiments of the application, also can use the accounting that account to be identified is used by the user corresponding with this account to be identified, this type of equivalent embodiment replaced all should be considered as within the protection domain of the application.In addition, above-mentioned active time can to represent in the use record of account to be identified the time interval between the use date to nearest use date the earliest usually, but, similarly, the application is not construed as limiting this, and it can also be the duration value of other feasible overall active time for representing account to be identified.

Generally speaking, in the embodiment of the present application, the active time of the account to be identified obtained in the second recognition unit 402 is longer, then can think that the credibility of account to be identified is higher, and judge account to be identified the probability that uses by user corresponding with it higher, otherwise, then judge account to be identified the probability that uses by user corresponding with it lower, on the other hand, if account to be identified is by not corresponding lower for the accounting used with it, then judge account to be identified the probability that uses by user corresponding with it higher, otherwise, then judge account to be identified the probability that uses by user corresponding with it lower, wherein, account to be identified is used to the judgement whether used by user corresponding with it in record in history, both can come from aforementioned first and identify the result judged, can also be known by other approach, such as come from the fed back statistics result out of the user corresponding with account to be identified.

Alternatively, in the embodiment of the present application, a time range can also be set for above-mentioned my statistics of use accounting non-, and in the second recognition unit 402, the non-service condition that I uses in this time range be added up.Such as, above-mentioned time range can for the date range before current date to N number of moon, thus in the second recognition unit 402, also can be that my access times non-are added up to the overall access times of account to be identified in this time range and the number of times used by the user not corresponding with this account to be identified, and use accounting using the ratio of non-access times and overall access times in person in this time range in person as non-.Under this scene, can add up for the use record that account to be identified is nearest, and get rid of the impact of the service condition of a period of time account to be identified in the past, thus the current service condition more adequately can treating identification account judges, obtain that there is more excellent ageing judged result to identify in judgement second.

More specifically, as shown in Figure 5, as the optional embodiment of one, the second recognition unit 402 can comprise:

1) the second acquisition module 502, enliven grade for what obtain each account to be identified according to the active time of each account to be identified, and obtained the accounting grade of each account to be identified by the accounting that the user not corresponding with account to be identified uses according to each account to be identified;

2) the second judge module 504, for judging the depth recognition grade of each account to be identified in each distribution situation enlivened in grade and each accounting grade according to all accounts to be identified, wherein, the depth recognition grade of account to be identified represents the relative probability that account to be identified is used by the user corresponding with account to be identified.

Wherein, be similar to the criterion identifying the range identification grade in judging aforementioned first, can also after the active time obtaining multiple account to be identified and accounting, according to the distribution situation of the account to be identified in the different quantized interval of active time and accounting, the credibility of the account to be identified of one of the quantized interval being positioned at above-mentioned active time and accounting is demarcated, such as, the account to be identified that can be in a certain quantized interval for active time and accounting is defined by the relative probability that user corresponding with it uses, and then can according to above-mentioned distribution situation or calibration result, the depth recognition grade judging account to be identified in judgement is identified second, wherein, namely this depth recognition grade may be used for the relative probability representing that this account to be identified is used by user corresponding with it.

Particularly, the quantized interval of above-mentioned active time and accounting can by being respectively active time and accounting divided rank realizes, dividing mode such as shown in table 5 and table 6, wherein, table 5 for represent mark off according to active time T enliven grade, table 6 is for representing the accounting grade marked off according to accounting P.

Table 5

The active time T(month)	Enliven grade L _T
		T<3	1
3≤T<12	2
		T≥12	3

Table 6

Accounting P(%)	Accounting grade L _P
		P<33	1
33≤P<66	2
		66≤P≤100	3

As shown in table 5, in the embodiment of the present application, can to enlivening grade L _tdo following division: by the active time section definition of 0 to 3 month for enlivening grade L _tequaling 1, is L by the active time section definition of 3 to 12 months _tequaling 2, is L by the active time section definition of more than 12 months _tequal 3, thus can realize obtaining each account to be identified according to the active time of each account to be identified in the second acquisition module 502 enliven grade, such as, if the active time of a certain account to be identified is 6 months, then what can obtain this account to be identified enlivens grade L _tequal 2.

Similarly, as shown in table 6, in the embodiment of the present application, can to accounting grade L _pdo following division: be accounting grade L by the accounting section definition of 0 to 33% _pequaling 1, is L by the active time section definition of 3 to 12 months _tequaling 2, is L by the active time section definition of more than 12 months _tequal 3, thus can realize obtaining each account to be identified according to the active time of each account to be identified in the second acquisition module 502 enliven grade, such as, if the active time of a certain account to be identified is 6 months, then what can obtain this account to be identified enlivens grade L _tequal 2.

Under above-mentioned scene, generally speaking, there is the higher account to be identified enlivening grade and lower accounting grade and usually can be considered as that there is higher credibility, thus can judge that the probability that account to be identified is used by user corresponding is with it higher, otherwise then can judge that probability is lower.But, in some embodiments of the application, be adopted as the deviation that mode that each combination enlivening grade and accounting grade gives the constant probable value of rigidity still may affect judged result and actual probabilities, thus in the second judge module 504, multiple account to be identified can be counted further in each distribution situation enlivened in grade and each accounting grade, and then judge the depth recognition grade of each account to be identified according to this distribution situation.

Be described in detail below in conjunction with the deterministic process of table 7 to the depth recognition grade of account to be identified.Wherein, table 7 is for representing the different quantity enlivening the account to be identified in the current account to be identified or historical record that grade and accounting grade distribute.

Table 7

In table 7, above-mentioned account to be identified can show as the quantity of the account to be identified in the combination that each enlivens grade and accounting grade, such as, for L in each distribution situation enlivened in grade and each accounting grade _t=1 and be L _p=2, also namely active time within 3 months and non-I use the combination of accounting between 33-66%, can for there are 50 accounts to be identified in the combination that this enlivens grade and accounting grade in statistics.

On the basis of the above, and then can according to the distribution situation in table 7, judge by the second judge module 502 relative probability that account to be identified is used by user corresponding with it, also the depth recognition grade of i.e. this account to be identified.Such as, consider that the usage behavior of most of user is optimum, thus according to statistical law, it can be grade combination imparting forward adjustment factor that the distributed quantity of the user to be identified in table 7 is relatively high, and be the relatively low grade combination imparting negative regulation coefficient of the distributed quantity of user to be identified, wherein, forward and negative regulation coefficient can represent the probability used by user corresponding with it for improving and reduce the user to be identified that judges respectively, and then obtain user to be identified by the relative probability of user corresponding with it use according to above-mentioned adjustment factor, also be depth recognition grade.

Certainly, above-described embodiment is the optional embodiment of one of technical scheme, and should not be construed as and made any restriction to the application.Such as, above-mentioned second identifies that judgement can also realize in the following manner, and wherein, as shown in Figure 6, the second acquisition module 502 can comprise:

1) second submodule 602 is obtained, grade is enlivened for what obtain each account to be identified according to the ratio between active time the longest in the active time of each account to be identified and all accounts to be identified, wherein, the difference belonged between each number enlivening the account to be identified of grade is less than predetermined threshold; And/or,

2) the 3rd submodule 604 is obtained, for obtaining the accounting grade of each account to be identified according to the ratio between accounting maximum in the accounting of each account to be identified and all accounts to be identified, wherein, the difference between the number belonging to the account to be identified of each accounting grade is less than predetermined threshold.

In the embodiment of the present application, obtain submodule 602 by the second acquisition submodule 602 and the 3rd and provide the dynamic dividing mode enlivening grade and accounting grade, wherein, different from the static division mode enlivening grade and accounting grade as shown in table 5 and table 6, in this dynamic dividing mode, the grade of enlivening of this account to be identified can be obtained by the ratio between the longest active time in the active time of the active time of a certain account to be identified and all accounts to be identified, such as shown in table 8, if under a scene, what in the account to be identified counted, active time was the longest is 60 months, thus, can obtain in submodule 602 second, obtain active time and compare R=T/60, such as, if T=30, also namely the active time of account to be identified is 30 months, then can show that the active time of this account to be identified compares R=50%, thus according to enlivening grade L shown in table 8 _tdivision rule, L can be obtained further _t=1, also namely the grade of enlivening of this account to be identified is the 1st grade.

Table 8

The active time T(month)	Active time is than R(%)	Enliven grade L _T	Account quantity to be identified
				T<20	R<33	1	200
20≤T<40	33≤R<66	2	100
				40≤T≤60	66≤R≤100	3	20

Further, in some embodiments of the application, the difference between each number enlivening the account to be identified of grade can also be may be limited in predetermined threshold.Such as, if the account multiple to be identified acquired is as shown in table 8 the 4th row, and above-mentioned predetermined threshold is set to 50, the number difference then enlivening the account to be identified of grade due to each has exceeded predetermined threshold, can be considered as current unreasonable to the dynamic dividing mode enlivening grade of user to be identified, and then can adjust this dynamic dividing mode, such as, can will be that 1 corresponding active time adjusts downwards than the upper limit judgment value 33% of R, to improve the accuracy of account identification further with enlivening grade.

Certainly, be more than a kind of example, the application is not limited in any way this, such as, also can show as other processing modes to the adjustment of dynamic dividing mode, such as can by the L in table 8 _t200 accounts to be identified of=1 do finer division, and different enliven grade for the different piece marked off is defined as.In fact, the optional embodiment enlivening grade obtaining account to be identified can also have multiple, and the application does not do tired stating at this.

Similarly, in above-mentioned dynamic dividing mode, can also obtain in submodule 604 the 3rd, the ratio between the maximum accounting in the accounting of accounting and all accounts to be identified is used to obtain the accounting grade of this account to be identified in person by the non-of a certain account to be identified, its concrete processing procedure is similar to the above-mentioned dynamic dividing mode enlivening grade, and the application does not repeat.

Certainly, in the embodiment of the present application, also can identify that by above-mentioned first judgement and second identifies that judgement combines, judge respectively account to be identified, such as, the higher distribution list of a dimension is set up in the two combination, and the application is not construed as limiting this.

On the basis of above-described embodiment, according to the account recognition system that the embodiment of the present application provides, in judging unit 104, can identify that the result and/or second judged identifies that the result judged judges whether account to be identified is used by the user corresponding with account to be identified according to first.

Generally speaking, in the applied environment of the account recognition system provided according to the embodiment of the present application, the judgement of final "Yes" or "No" made by usual needs, than online payment scene as the aforementioned, need a series of judged result of drawing according to recognition unit 102 or probability numbers, finally make whether by the decision of the payment request of account to be identified.For addressing this problem, in judging unit 104, can identify that judgement and/or second identifies that the result judged judges whether account to be identified is used by the user corresponding with account to be identified according to first, wherein, particularly, the threshold value of a certain probability or relative probability can be set for this judgement, when first identifies judgement or second identifies the result of judgement higher than this threshold value, judge that account to be identified is used by user corresponding with it, otherwise judge that account to be identified is not used by user corresponding with it.

Certainly, it is more than a kind of example, in the embodiment of the present application, the judged result drawn by judging unit 104 also can not be non-is namely no binaryzation result, also can for marking off the judged result of multiple grade, such as shown in Figure 14, in the embodiment of the present application, judging unit 104 can also comprise:

1) the 3rd judge module 1402, for the relative probability used according to range identification grade and the depth recognition grade acquisition user corresponding with account to be identified of each account to be identified.

In the embodiment of the present application, the concrete decision logic of the 3rd judge module 1402, also namely according to the range identification grade of account to be identified and depth recognition grade judge this account to be identified the mode of relative probability that uses by user corresponding with it can have multiple.Such as, as wherein a kind of feasible embodiment, the decision logic of the 3rd judge module 1402 can be similar to the mode judging the depth recognition grade of each account to be identified according to all accounts to be identified in each distribution situation enlivened in grade and each accounting grade described in the second judge module 504, also be, after the range identification grade judging account to be identified respectively by the first judge module 204 and the second judge module 504 and depth recognition grade, can utilize further the 3rd judge module 1402 according to all accounts to be identified the distribution situation in each range identification grade and each depth recognition grade judge account to be identified the probability that uses by user corresponding with it.

Certainly, in the embodiment of the present application, can also carry out other feasible process according to the range of the account to be identified obtained or depth recognition information, the application is not construed as limiting this.But should be understood that, this embodiment must be considered as within the protection domain of the application.

Embodiment 2

On the basis of the account recognition system shown in Fig. 1 to Fig. 6, the embodiment of the present application additionally provides a kind of account recognition methods.As shown in Figure 7, the method can comprise:

S702: the consistent chain relevant information of information obtaining the same attribute of account to be identified, and according to information consistent chain relevant information, the first identification judgement is carried out to account to be identified;

S704: identify that the result judged judges whether account to be identified is used by the user corresponding with account to be identified according to first.

First will be clear that, one of technical problems to be solved in this application, be to provide a kind of recognition methods, by this recognition methods, can to the account whether by the user corresponding with the account, the registered user of the such as account uses and judges, on the other hand, by the judged result that the method obtains, also can be used for reflecting the credibility of the account.Such as, this account recognition methods can be applied to the scene of network payment, wherein, when server receives the payment request of a certain account, the account recognition methods provided according to the embodiment of the present application can be utilized to judge the credibility of the account, whether to make the account further in the judgement used by user corresponding with it, and finally make by or refuse the decision-making of this payment request.For ease of describing, below the identification object of account recognition methods is designated as account to be identified.

For the purpose of more effectively solving the problem, in the account recognition methods that the embodiment of the present application provides, the concept of the consistent chain of the information that proposes.Particularly, for an account to be identified, the consistent chain of the information corresponding to it can comprise one or more index.In a consistent chain of information, one or more index should belong to above-mentioned same attribute, the content of these indexs should be consistent with each other, and the content of index should be corresponding with account to be identified the log-on message of user consistent, at least should comprise in the log-on message of the user corresponding with account to be identified the index belonging to this attribute in other words in these indexs.

The example provided below in conjunction with the table 1 shown in embodiment 1 is described in detail the consistent chain of information.Wherein, this account to be identified, for representing an account-related information record to be identified, for ease of describing, is designated as the first account by table 1 below.

It should be noted that, the relevant information record recorded in the process of the above-mentioned business of the first account process, usually the collection of information items be combined to form by more than one item of information can be shown as, such as, for certification or registration business, the collection of information items corresponding with the first account can comprise the item of information of same attribute usually: name, the pet name, age, identification card number, address, telephone number, email address etc., and in other business, the concrete form of the single item of information in the collection of information items corresponding with the first account both can be similar text message, also other data type and structure can be shown, such as the business of receiving, its information of receiving also can comprise individual subscriber and arrange the information such as code.But, in the collection of information items recorded when use first account process different business, usually there is one or more item of information belonging to same alike result, name as escribed above, or other feasible information content, such as phone numbers etc.

On basis described above, in the embodiment of the present application, the consistent chain relevant information of information of the same attribute of account to be identified can be obtained in step S702, and according to information consistent chain relevant information, the first identification judgement is carried out to account to be identified.Wherein, information consistent chain relevant information can comprise one of following:

As shown in Figure 8, as a kind of optional embodiment of the application, in step S702, according to information consistent chain relevant information, the first identification is carried out to account to be identified and judges to comprise:

S802: the attribute ratings corresponding according to each attribute of each account to be identified of information consistent chain relevant information acquisition, wherein, the attribute ratings that a certain attribute is corresponding represents the relative probability that the account to be identified that goes out based on this determined property is used by the user corresponding with account to be identified;

S804: the range identification grade judging each account to be identified according to the distribution situation of all accounts to be identified on each probability levels of each attribute, wherein, the range identification grade of account to be identified represents the relative probability that account to be identified is used by the user corresponding with account to be identified.

Particularly, above-mentioned grade interval can the different information divided rank by the information being respectively a certain attribute in consistent chain relevant information realize, such as, the dividing mode shown in table 2 in embodiment 1, wherein, table 2 is with representing the attribute ratings marked off according to the length of the consistent chain of the information of the name information in relative recording information.

As shown in table 2, in the embodiment of the present application, can to attribute ratings L _ado following division: the attribute ratings section definition being 1 to 3 by the length of consistent for information chain is attribute ratings L _aequaling 1, is that the attribute ratings section definition of 4 to 9 is L by length _aequaling 2, is that the attribute ratings section definition of more than 10 is L by length _aequal 3, thus the length that can realize the consistent chain of information of the name attribute according to each account to be identified in step S802 obtains the attribute ratings of each account to be identified, such as, if the length of the consistent chain of information of the name attribute of a certain account to be identified is 8, then can draw the attribute ratings L of this account to be identified _aequal 2.

Under above-mentioned scene, generally speaking, the account to be identified with higher attribute ratings can be considered as having higher credibility usually, thus can judge that the probability that account to be identified is used by user corresponding is with it higher, otherwise then can judge that probability is lower.But, in some embodiments of the application, be adopted as the deviation that mode that each attribute ratings gives the constant probable value of rigidity still may affect judged result and actual probabilities, thus in step S804, the distribution situation of multiple accounts to be identified in each attribute ratings can be counted further, and then the depth recognition grade of each account to be identified can be judged according to this distribution situation.

Be described in detail below in conjunction with the deterministic process of the table 3 shown in embodiment 1 to the range identification grade of account to be identified.Wherein, table 3 is for representing the quantity of the account to be identified that different attribute ratings distributes.

On the basis of the above, and then can according to the distribution situation in table 3, judge in step S804 the relative probability that account to be identified is used by user corresponding with it, also the range identification grade of i.e. this account to be identified.Such as, consider that the usage behavior of most of user is optimum, thus according to statistical law, it can be grade combination imparting forward adjustment factor that the distributed quantity of the user to be identified in table 3 is relatively high, and be the relatively low grade combination imparting negative regulation coefficient of the distributed quantity of user to be identified, wherein, forward and negative regulation coefficient can represent the probability used by user corresponding with it for improving and reduce the user to be identified that judges respectively, and then obtain user to be identified by the relative probability of user corresponding with it use according to above-mentioned adjustment factor, also be range identification grade.

Further, in some embodiments of the application, range identification grade can also judge in conjunction with the attribute ratings of multiple attribute, shown in such as, table 4 in embodiment 1, and wherein L _a-Nthe attribute ratings that " name " attribute of drawing according to the length of the consistent chain of information is corresponding can be represented, and L _a-Ithe attribute ratings that " identification card number " attribute of drawing according to the length of the consistent chain of information is corresponding can be represented.

Certainly, it is more than a kind of example, in the embodiment of the present application, the decision logic of range identification grade can also have multiple, such as, can also respectively according to the length of the consistent chain of information, the number of predetermined important attribute in the consistent chain of information or ratio, and/or the same index of the consistent chain of information judges the range identification grade of account to be identified, and then, the range identification grade of account to be identified can also be judged according to the distribution situation in each combination of multiple attribute ratings of user to be identified between above-mentioned different information consistent chain relevant information, the application does not do tired stating at this.

And then as shown in Figure 9, in the embodiment of the present application, step S802 can also comprise:

S902: the attribute ratings corresponding according to the first attribute of each account to be identified of distributed acquisition of the length of the consistent chain of information of the first attribute of all accounts to be identified, wherein,

On the other hand, in the embodiment of the present application, except can carry out the first identification judgement in step S702 and in step S704 according to first identify the result judged judge account to be identified whether use by the user corresponding with this account to be identified except, alternatively, as shown in Figure 10, above-mentioned account recognition methods can also comprise:

S1002: obtain the accounting that the active time of account to be identified and account to be identified are used by the user not corresponding with account to be identified, and according to active time and accounting, the second identification judgement is carried out to account to be identified; Wherein,

Step S704 can comprise:

S1004: identify that the result and second judged identifies that the result judged judges whether account to be identified is used by the user corresponding with account to be identified according to first.

Generally speaking, in the embodiment of the present application, the active time of the account to be identified obtained in step S1002 is longer, then can think that the credibility of account to be identified is higher, and judge account to be identified the probability that uses by user corresponding with it higher, otherwise, then judge account to be identified the probability that uses by user corresponding with it lower, on the other hand, if account to be identified is by not corresponding lower for the accounting used with it, then judge account to be identified the probability that uses by user corresponding with it higher, otherwise, then judge account to be identified the probability that uses by user corresponding with it lower, wherein, account to be identified is used to the judgement whether used by user corresponding with it in record in history, both can come from aforementioned first and identify the result judged, can also be known by other approach, such as come from the fed back statistics result out of the user corresponding with account to be identified.

Alternatively, in the embodiment of the present application, a time range can also be set for above-mentioned my statistics of use accounting non-, and in step S1002, the non-service condition that I uses in this time range be added up.Such as, above-mentioned time range can for the date range before current date to N number of moon, thus in step S1002, also can be that my access times non-are added up to the overall access times of account to be identified in this time range and the number of times used by the user not corresponding with this account to be identified, and use accounting using the ratio of non-access times and overall access times in person in this time range in person as non-.Under this scene, can add up for the use record that account to be identified is nearest, and get rid of the impact of the service condition of a period of time account to be identified in the past, thus the current service condition more adequately can treating identification account judges, obtain that there is more excellent ageing judged result to identify in judgement second.

More specifically, as shown in figure 11, as the optional embodiment of one, in step S1002, according to active time and accounting, the second identification is carried out to account to be identified and judges to comprise:

S1102: what obtain each account to be identified according to the active time of each account to be identified enlivens grade, and is obtained the accounting grade of each account to be identified by the accounting that the user not corresponding with account to be identified uses according to each account to be identified;

S1104: the depth recognition grade judging each account to be identified according to all accounts to be identified in each distribution situation enlivened in grade and each accounting grade, wherein, the depth recognition grade of account to be identified represents the relative probability that account to be identified is used by the user corresponding with account to be identified.

Particularly, the quantized interval of above-mentioned active time and accounting can by being respectively active time and accounting divided rank realizes, such as, table 5 in embodiment 1 and the dividing mode shown in table 6, wherein, table 5 for represent mark off according to active time T enliven grade, table 6 is for representing the accounting grade marked off according to accounting P.

As shown in table 5, in the embodiment of the present application, can to enlivening grade L _tdo following division: by the active time section definition of 0 to 3 month for enlivening grade L _tequaling 1, is L by the active time section definition of 3 to 12 months _tequaling 2, is L by the active time section definition of more than 12 months _tequal 3, thus can realize obtaining each account to be identified according to the active time of each account to be identified in step S1102 enliven grade, such as, if the active time of a certain account to be identified is 6 months, then what can obtain this account to be identified enlivens grade L _tequal 2.

Similarly, as shown in table 6, in the embodiment of the present application, can to accounting grade L _pdo following division: be accounting grade L by the accounting section definition of 0 to 33% _pequaling 1, is L by the active time section definition of 3 to 12 months _tequaling 2, is L by the active time section definition of more than 12 months _tequal 3, thus can realize obtaining each account to be identified according to the active time of each account to be identified in step S1102 enliven grade, such as, if the active time of a certain account to be identified is 6 months, then what can obtain this account to be identified enlivens grade L _tequal 2.

Under above-mentioned scene, generally speaking, there is the higher account to be identified enlivening grade and lower accounting grade and usually can be considered as that there is higher credibility, thus can judge that the probability that account to be identified is used by user corresponding is with it higher, otherwise then can judge that probability is lower.But, in some embodiments of the application, be adopted as the deviation that mode that each combination enlivening grade and accounting grade gives the constant probable value of rigidity still may affect judged result and actual probabilities, thus in step S1104, multiple account to be identified can be counted further in each distribution situation enlivened in grade and each accounting grade, and then judge the depth recognition grade of each account to be identified according to this distribution situation.

Be described in detail below in conjunction with the deterministic process of the table 7 shown in embodiment 1 to the depth recognition grade of account to be identified.Wherein, table 7 is for representing the different quantity enlivening the account to be identified in the current account to be identified or historical record that grade and accounting grade distribute.

On the basis of the above, and then can according to the distribution situation in table 7, judge in step S1104 the relative probability that account to be identified is used by user corresponding with it, also the depth recognition grade of i.e. this account to be identified.Such as, consider that the usage behavior of most of user is optimum, thus according to statistical law, it can be grade combination imparting forward adjustment factor that the distributed quantity of the user to be identified in table 7 is relatively high, and be the relatively low grade combination imparting negative regulation coefficient of the distributed quantity of user to be identified, wherein, forward and negative regulation coefficient can represent the probability used by user corresponding with it for improving and reduce the user to be identified that judges respectively, and then obtain user to be identified by the relative probability of user corresponding with it use according to above-mentioned adjustment factor, also be depth recognition grade.

Certainly, above-described embodiment is the optional embodiment of one of technical scheme, and should not be construed as and made any restriction to the application.Such as, above-mentioned second identifies that judgement can also realize in the following manner, and wherein, as shown in figure 12, step S1104 can comprise:

S1202: what obtain each account to be identified according to the ratio between active time the longest in the active time of each account to be identified and all accounts to be identified enlivens grade, wherein, the difference belonged between each number enlivening the account to be identified of grade is less than predetermined threshold; And/or,

S1204: the accounting grade obtaining each account to be identified according to the ratio between accounting maximum in the accounting of each account to be identified and all accounts to be identified, wherein, the difference between the number belonging to the account to be identified of each accounting grade is less than predetermined threshold.

In the embodiment of the present application, the dynamic dividing mode enlivening grade and accounting grade is provided by step S1202 and S1204, wherein, different from the static division mode enlivening grade and accounting grade as shown in table 5 and table 6, in this dynamic dividing mode, the grade of enlivening of this account to be identified can be obtained by the ratio between the longest active time in the active time of the active time of a certain account to be identified and all accounts to be identified, shown in such as, table 8 in embodiment 1, if under a scene, what in the account to be identified counted, active time was the longest is 60 months, thus, can in step S1202, obtain active time and compare R=T/60, such as, if T=30, also namely the active time of account to be identified is 30 months, then can show that the active time of this account to be identified compares R=50%, thus according to enlivening grade L shown in table 8 _tdivision rule, L can be obtained further _t=1, also namely the grade of enlivening of this account to be identified is the 1st grade.

Similarly, in above-mentioned dynamic dividing mode, can also in step S1204, the ratio between the maximum accounting in the accounting of accounting and all accounts to be identified is used to obtain the accounting grade of this account to be identified in person by the non-of a certain account to be identified, its concrete processing procedure is similar to the above-mentioned dynamic dividing mode enlivening grade, and the application does not repeat.

On the basis of above-described embodiment, according to the account recognition methods that the embodiment of the present application provides, in step S704, can identify that the result and/or second judged identifies that the result judged judges whether account to be identified is used by the user corresponding with account to be identified according to first.

Generally speaking, in the applied environment of the account recognition methods provided according to the embodiment of the present application, the judgement of final "Yes" or "No" made by usual needs, than online payment scene as the aforementioned, need a series of judged result of drawing according to step S702 or probability numbers, finally make whether by the decision of the payment request of account to be identified.For addressing this problem, in step S704, can identify that judgement and/or second identifies that the result judged judges whether account to be identified is used by the user corresponding with account to be identified according to first, wherein, particularly, the threshold value of a certain probability or relative probability can be set for this judgement, when first identifies judgement or second identifies the result of judgement higher than this threshold value, judge that account to be identified is used by user corresponding with it, otherwise judge that account to be identified is not used by user corresponding with it.

Certainly, be more than a kind of example, in the embodiment of the present application, the judged result that step S704 draws also can not for non-be namely no binaryzation result, also can for marking off the judged result of multiple grade, such as, shown in Figure 13, in the embodiment of the present application, step S1004 can comprise:

S1302: the relative probability used according to range identification grade and the depth recognition grade acquisition user corresponding with account to be identified of each account to be identified.

In the embodiment of the present application, the range identification grade according to account to be identified described in step S1302 and depth recognition grade judge this account to be identified the embodiment of relative probability that uses by user corresponding with it can have multiple.Such as, as wherein a kind of feasible mode, the processing logic of step S1302 can be similar to the mode judging the depth recognition grade of each account to be identified according to all accounts to be identified in each distribution situation enlivened in grade and each accounting grade described in step S1104, also be, after the range identification grade judging account to be identified respectively by step S804 and S1104 and depth recognition grade, can further in step S1302 according to all accounts to be identified the distribution situation in each range identification grade and each depth recognition grade judge account to be identified the probability that uses by user corresponding with it.

Still it should be noted that, those skilled in the art should be understood that, each module of above-mentioned the application or each step can realize with general calculation element, they can concentrate on single calculation element, or be distributed on network that multiple calculation element forms, alternatively, they can realize with the executable program code of calculation element, thus, they can be stored and be performed by calculation element in the storage device, and in some cases, step shown or described by can performing with the order be different from herein, or they are made into each integrated circuit modules respectively, or the multiple module in them or step are made into single integrated circuit module to realize.Like this, the application is not restricted to any specific hardware and software combination.

The foregoing is only the preferred embodiment of the application, be not limited to the application, for a person skilled in the art, the application can have various modifications and variations.Within all spirit in the application and principle, any amendment done, equivalent replacement, improvement etc., within the protection domain that all should be included in the application.

Claims

1. an account recognition methods, is characterized in that, comprising:

Obtain the consistent chain relevant information of information of the same attribute of account to be identified, and according to described information consistent chain relevant information, the first identification judgement is carried out to described account to be identified;

At least identify that the result judged judges whether described account to be identified is used by the user corresponding with described account to be identified according to described first;

Wherein, described information consistent chain relevant information comprises one of following:

The length of the consistent chain of information, wherein, the consistent chain of described information comprises and belongs to described same attribute and the log-on message of the content user corresponding with described account to be identified one or more indexs consistent with each other, and the length of the consistent chain of described information represents the number of the described index that the consistent chain of described information comprises;

The number of predetermined important indicator is comprised in the length chain consistent with described information of the consistent chain of described information; Or

The same index that the number chain consistent with described information of predetermined important indicator is corresponding is comprised in the length of the consistent chain of described information, the consistent chain of described information.

2. method according to claim 1, is characterized in that, described attribute comprises multiple, wherein, described according to described information consistent chain relevant information to described account to be identified carry out first identification judge comprise:

The attribute ratings corresponding according to each attribute of each described account to be identified of described information consistent chain relevant information acquisition, wherein, the attribute ratings that described attribute is corresponding represents the relative probability that the account described to be identified that goes out based on described determined property is used by the user corresponding with described account to be identified;

The range identification grade of each described account to be identified is judged according to the distribution situation of all described accounts to be identified on each probability levels of each described attribute, wherein, the range identification grade of described account to be identified represents the relative probability that described account to be identified is used by the user corresponding with described account to be identified.

3. method according to claim 2, is characterized in that, described each attribute comprises the first attribute, and wherein, the described attribute ratings corresponding according to each attribute of each described account to be identified of described information consistent chain relevant information acquisition comprises:

The attribute ratings corresponding according to described first attribute of each described account to be identified of the distributed acquisition of the length of the consistent chain of information of the first attribute of all described accounts to be identified, wherein,

For the first account to be identified that the length of the consistent chain of information described in all described accounts to be identified is identical, the attribute ratings corresponding according to described first attribute of described first account to be identified of number acquisition comprising predetermined important indicator in the distribution of the length of the consistent chain of information of the first attribute of described first account to be identified and the consistent chain of information of described first attribute of described first account to be identified;

For the length of the consistent chain of information described in all described accounts to be identified identical and described in comprise the second identical account to be identified of the number of predetermined important indicator, the same index corresponding according to the number and the consistent chain of described information that comprise predetermined important indicator in the distribution of the length of the consistent chain of information of the first attribute of described second account to be identified and the consistent chain of information of described first attribute of described second account to be identified obtains attribute ratings corresponding to described first attribute of described second account to be identified.

4. method according to claim 2, is characterized in that, described at least identify that the result judged judges that whether described account to be identified is used by the user corresponding with described account to be identified according to described first before, also comprise:

Obtain the accounting that the active time of described account to be identified and described account to be identified are used by the user not corresponding with described account to be identified, and according to described active time and described accounting, the second identification judgement is carried out to described account to be identified; Wherein,

Describedly at least identify that the result judged judges that whether described account to be identified is used by the user corresponding with described account to be identified and comprises according to described first: identify that the result and described second judged identifies that the result judged judges whether described account to be identified is used by the user corresponding with described account to be identified according to described first.

5. method according to claim 4, is characterized in that, describedly carries out the second identification according to described active time and described accounting to described account to be identified and judges to comprise:

What obtain each described account to be identified according to the active time of each described account to be identified enlivens grade, and is obtained the accounting grade of each described account to be identified by the accounting that the user not corresponding with described account to be identified uses according to each described account to be identified;

Judge the depth recognition grade of each described account to be identified in each described distribution situation enlivened in grade and described each accounting grade according to all described accounts to be identified, wherein, the depth recognition grade of described account to be identified represents the relative probability that described account to be identified is used by the user corresponding with described account to be identified.

6. method according to claim 5, is characterized in that, the grade of enlivening that the described active time according to each described account to be identified obtains each described account to be identified comprises:

What obtain each described account to be identified according to the ratio between active time the longest in the active time of each described account to be identified and all described accounts to be identified enlivens grade, wherein, the difference belonged between each described number enlivening the account described to be identified of grade is less than predetermined threshold.

7. method according to claim 5, is characterized in that, is describedly comprised by the accounting grade that the accounting that the user not corresponding with described account to be identified uses obtains each described account to be identified according to each described account to be identified:

The accounting grade of each described account to be identified is obtained according to the ratio between accounting maximum in the described accounting of each described account to be identified and all described accounts to be identified, wherein, the difference between the number belonging to the account described to be identified of each described accounting grade is less than predetermined threshold.

8. method according to claim 5, is characterized in that, describedly identifies that the result and described second judged identifies that the result judged judges that whether described account to be identified is used by the user corresponding with described account to be identified and comprises according to described first:

According to the relative probability that range identification grade and the depth recognition grade acquisition user corresponding with described account to be identified of each described account to be identified use.

9. an account recognition system, is characterized in that, comprising:

First recognition unit, the consistent chain relevant information of the information for the same attribute obtaining account to be identified, and according to described information consistent chain relevant information, the first identification judgement is carried out to described account to be identified;

According to described first, judging unit, at least identifying that the result judged judges whether described account to be identified is used by the user corresponding with described account to be identified;

10. system according to claim 9, is characterized in that, described attribute comprises multiple, and wherein, described first recognition unit comprises:

First acquisition module, for the attribute ratings that each attribute obtaining each described account to be identified according to described information consistent chain relevant information is corresponding, wherein, the attribute ratings that described attribute is corresponding represents the relative probability that the account described to be identified that goes out based on described determined property is used by the user corresponding with described account to be identified;

First judge module, for judging the range identification grade of each described account to be identified according to the distribution situation of all described accounts to be identified on each probability levels of each described attribute, wherein, the range identification grade of described account to be identified represents the relative probability that described account to be identified is used by the user corresponding with described account to be identified.

11. systems according to claim 10, is characterized in that, described each attribute comprises the first attribute, and wherein, described first acquisition module comprises:

First obtains submodule, the attribute ratings that described first attribute of each described account to be identified of the distributed acquisition for the length of the consistent chain of information of the first attribute according to all described accounts to be identified is corresponding, wherein,

12. systems according to claim 10, is characterized in that, also comprise:

Second recognition unit, for the accounting that the active time and described account to be identified that obtain described account to be identified are used by the user not corresponding with described account to be identified, and according to described active time and described accounting, the second identification judgement is carried out to described account to be identified; Wherein,

Described judging unit is used for identifying that the result and described second judged identifies that the result judged judges whether described account to be identified is used by the user corresponding with described account to be identified according to described first.

13. systems according to claim 12, is characterized in that, described second recognition unit comprises:

Second acquisition module, enliven grade for what obtain each described account to be identified according to the active time of each described account to be identified, and obtained the accounting grade of each described account to be identified by the accounting that the user not corresponding with described account to be identified uses according to each described account to be identified;

Second judge module, for judging the depth recognition grade of each described account to be identified in each described distribution situation enlivened in grade and described each accounting grade according to all described accounts to be identified, wherein, the depth recognition grade of described account to be identified represents the relative probability that described account to be identified is used by the user corresponding with described account to be identified.

14. systems according to claim 13, is characterized in that, described second acquisition module comprises:

Second obtains submodule, grade is enlivened for what obtain each described account to be identified according to the ratio between active time the longest in the active time of each described account to be identified and all described accounts to be identified, wherein, the difference belonged between each described number enlivening the account described to be identified of grade is less than predetermined threshold.

15. systems according to claim 13, is characterized in that, described second acquisition module comprises:

3rd obtains submodule, for obtaining the accounting grade of each described account to be identified according to the ratio between accounting maximum in the described accounting of each described account to be identified and all described accounts to be identified, wherein, the difference between the number belonging to the account described to be identified of each described accounting grade is less than predetermined threshold.

16. systems according to claim 13, is characterized in that, described judging unit comprises:

3rd judge module, for the relative probability used according to range identification grade and the depth recognition grade acquisition user corresponding with described account to be identified of each described account to be identified.