CN106919816A

CN106919816A - A kind of user authen method and device, a kind of device for user authentication

Info

Publication number: CN106919816A
Application number: CN201510993000.0A
Authority: CN
Inventors: 崔欣; 张扬
Original assignee: Beijing Sogou Technology Development Co Ltd
Current assignee: Beijing Sogou Technology Development Co Ltd
Priority date: 2015-12-24
Filing date: 2015-12-24
Publication date: 2017-07-04

Abstract

A kind of user authen method, a kind of user authentication device and a kind of device for user authentication are the embodiment of the invention provides, method therein is specifically included：The input behavior feature of the cryptogram of collection active user's input；Cryptogram to active user input carries out the first certification；After first certification passes through, second certification is carried out to the input behavior feature according to the corresponding characteristic model of the input behavior feature, to obtain corresponding second authentication result, when second certification passes through, determine active user's certification success, it is obstructed out-of-date in second certification, determine active user's authentification failure；Wherein, the characteristic model is the model that the history input behavior feature according to the validated user of the cryptogram corresponding account when the cryptogram or cryptogram correspondence cryptogram fragment is input into is set up.The embodiment of the present invention can improve the security of user authentication.

Description

A kind of user authen method and device, a kind of device for user authentication

Technical field

The present invention relates to security technology area, more particularly to a kind of user authen method, a kind of user Authentication device and a kind of device for user authentication.

Background technology

In recent years, the various information equipments with computer as representative are popularized.Particularly there is electricity The mobile terminal of sub- mail function and internet access function obtains rapidly popularization, as most of The essential information tool of user.

Along with the development of informationized society, the safety problems such as the abnormal access of system are become particularly It is important.In order to prevent the unauthorized access to system, existing user authentication scheme generally uses login in advance ID and password carry out user authentication, namely user input ID and password, advance by user The ID and password of setting compare completion user authentication.

However, due to the input of password be user directly on keyboard by it is corresponding numeral or letter key Complete, the third party need to only record or watch the input process of the password, or using cracking time Going through trial can just steal the password.In addition, the password being input into Network is also easy in transmission During be intercepted.Therefore, the characteristic for being easily stolen and intercepting that the password has, causes existing There is the security of user's certificate scheme not high.

The content of the invention

In view of the above problems, it is proposed that the embodiment of the present invention so as to provide one kind overcome above mentioned problem or The user authen method and user authentication device that solve the above problems at least in part and for user authentication Device, it is possible to increase the security of user authentication.

In order to solve the above problems, the invention discloses a kind of user authen method, including：

The input behavior feature of the cryptogram of collection active user's input；

Cryptogram to active user input carries out the first certification；

After first certification passes through, according to the corresponding characteristic model of the input behavior feature to described Input behavior feature carries out the second certification, to obtain corresponding second authentication result, in second certification By when, determine active user's certification success, it is obstructed out-of-date in second certification, it is determined that current use Family authentification failure；Wherein, the characteristic model is the legal use according to the cryptogram corresponding account History input of the family when the cryptogram or cryptogram correspondence cryptogram fragment is input into The model that behavioural characteristic is set up.

Preferably, the characteristic model is built for an input behavior feature according to the cryptogram Vertical model, then it is described special to the input behavior according to the corresponding characteristic model of the input behavior feature The step of carrying out the second certification is levied, including：

The input behavior feature is input into corresponding characteristic model, exports right by the characteristic model The first confidence level answered；

According to corresponding first confidence level of all input behavior features, the confidence level of active user is obtained；

According to the confidence level of active user, judge whether second certification passes through.

Preferably, the input model is built for the multiple input behavior features according to the cryptogram Vertical model, then it is described special to the input behavior according to the corresponding characteristic model of the input behavior feature The step of carrying out the second certification is levied, including：

Multiple input behavior features are input into corresponding characteristic model, are worked as by characteristic model output The confidence level of preceding user；

Preferably, it is described according to the corresponding characteristic model of the input behavior feature to the input behavior The step of feature carries out the second certification, including：

The cryptogram is cut into multiple cryptogram fragments；Wherein, the cryptogram fragment tool There is corresponding input behavior characteristic fragment；

The corresponding input behavior characteristic fragment of the cryptogram fragment is input into corresponding characteristic model In, corresponding second confidence level is exported by the characteristic model；

According to corresponding second confidence level of all cryptogram fragments, the confidence level of active user is obtained；

Preferably, the length of each cryptogram fragment is identical, or, the length of each cryptogram fragment Degree is different；

Wherein, the length of each cryptogram fragment is different, including：The cryptogram fragment Length according to it is described it is vertical sequentially.

Preferably, it is described according to the corresponding characteristic model of the input behavior feature to the line of input Before being characterized the step of carrying out the second certification, methods described also includes：

Search the index with the cryptogram as major key；Wherein, the index is with the cryptogram Major key, the storage location with the characteristic model is as storage value；

According to the storage location for indexing and obtaining characteristic model, and obtain corresponding characteristic model.

Preferably, methods described also includes：

When the index with the cryptogram as major key is not found, search with the drop of the cryptogram Rank text is the first index of major key；

The storage location of characteristic model is obtained according to the described first index, and obtains corresponding characteristic model.

Preferably, it is characterised in that the characteristic model is gauss hybrid models, then as follows Set up the characteristic model：

Collect the corresponding history input behavior feature of cryptogram, and/or, cryptogram fragment is corresponding goes through History input behavior characteristic fragment；

Calculate the history input behavior feature and/or the history input behavior characteristic fragment average and Variance, and set up the gauss hybrid models with the average and variance as parameter.

Preferably, methods described also includes：

When second certification passes through, according to the input of the active user correspondence cryptogram Behavioural characteristic carries out the first renewal to the corresponding characteristic model of the input behavior feature, or, according to close The input behavior characteristic fragment of code text fragments is to the corresponding characteristic model of the input behavior characteristic fragment Carry out the second renewal.

Preferably, the characteristic model is gauss hybrid models, then the step of first renewal, bag Include：

Obtain the cryptogram history input number of times, the first average of history input behavior feature and First variance, and according to the input behavior feature of the active user correspondence cryptogram, to described First average and first variance are updated；

The step of second renewal, including：

Obtain history input number of times, the of history input behavior characteristic fragment of the cryptogram fragment Two averages and second variance, and according to the input behavior of the active user correspondence cryptogram fragment Characteristic fragment, is updated to second average and second variance.

Preferably, the input behavior feature includes：In key press time, button coordinate and key press At least one.

On the other hand, the invention also discloses a kind of user authentication device, including：

Acquisition module, the input behavior feature of the cryptogram for gathering active user's input；

First authentication module, for carrying out the first certification to the cryptogram that the active user is input into；

Second authentication module, for after first certification passes through, according to the input behavior feature pair The characteristic model answered carries out the second certification to the input behavior feature, to obtain corresponding second certification As a result, when second certification passes through, active user's certification success is determined, in second certification not By when, determine active user's authentification failure；Wherein, the characteristic model is according to password text The validated user of this corresponding account is input into the cryptogram or cryptogram correspondence password text The model that history input behavior feature during this fragment is set up.

Another further aspect, the invention also discloses a kind of device for user authentication, including：Memory, And one or more than one program, one of them or more than one program storage is in memory In, and be configured to by one or more than one computing device is one or more than one journey Sequence includes the instruction for carrying out following operation：

Cryptogram to active user input carries out the first certification；

The embodiment of the present invention includes advantages below：

The embodiment of the present invention can be according to active user when the input of cryptogram is carried out input behavior Feature to user carries out user authentication, is specifically as follows the user for passing through the first certification to cryptogram, Second certification is carried out according to the corresponding characteristic model of cryptogram；Features described above model is specifically as follows basis The spy that the history input behavior data of the validated user above-mentioned cryptogram of correspondence of above-mentioned cryptogram are set up Model is levied, when being input into identical cryptogram due to different user, corresponding input behavior feature is often It is the presence of larger difference, and above-mentioned input behavior characteristic model can reflect in validated user input State the rule of the input behavior feature of cryptogram；So, illegally used in the cryptogram of validated user After family is stolen, the input of input behavior feature when inputing the password due to disabled user and validated user There is larger difference in behavioural characteristic, this causes that the input behavior feature of disabled user does not meet above-mentioned spy The rule of the input behavior feature of model reaction is levied, and then causes above-mentioned disabled user from by the Two certifications, therefore, the embodiment of the present invention can improve the security of user authentication.

Brief description of the drawings

The step of Fig. 1 is a kind of user authen method embodiment one of the invention flow chart；

Fig. 2 is a kind of schematic diagram of key area distribution of the invention；

Fig. 3 be it is provided in an embodiment of the present invention it is a kind of set up characteristic model method the step of flow chart；

The step of Fig. 4 is a kind of user authen method embodiment two of the invention flow chart；

The step of Fig. 5 is a kind of user authen method embodiment three of the invention flow chart；

The step of Fig. 6 is a kind of user authen method example IV of the invention flow chart；

The step of Fig. 7 is a kind of user authen method example of present invention offer flow chart；

The step of Fig. 8 is a kind of user authen method embodiment five of the invention flow chart；

Fig. 9 is a kind of structured flowchart of user authentication device embodiment of the invention；

Figure 10 is a kind of block diagram of device 1000 for recommending of the invention；And

Figure 11 is a kind of structural representation of server of the invention.

Specific embodiment

To enable the above objects, features and advantages of the present invention more obvious understandable, with reference to attached The present invention is further detailed explanation with specific embodiment for figure.

Password of the existing user authentication process often only to active user's input is authenticated.Illegally using When family has been stolen the password of validated user and has been input into, the password due to system only to being input into is carried out in itself Certification so that disabled user can pretend to be validated user, and then to the privacy and property safety of validated user Threaten.

And the embodiment of the present invention finds, when user carries out the input of cryptogram, often with personalization Input behavior feature, above-mentioned input behavior feature can specifically include：Key press time, button coordinate, presses Key pressure etc.；That is, the receipts of input behavior feature when cryptogram is input into by the history to user Collection and accumulation analysis, the input behavior feature of user's correspondence cryptogram can often be followed with regular；This Sample, can combine the input behavior feature of the user for currently carrying out authentication, and the active user is carried out Authentication, improves the security of authenticating user identification.

Therefore, the embodiment of the present invention creatively proposes, can carry out password text according to active user Input behavior feature during this input carries out user authentication to user, is specifically as follows to password text This carries out the second certification by the user of the first certification according to the corresponding characteristic model of cryptogram；It is above-mentioned Characteristic model is specifically as follows going through according to the validated user of the above-mentioned cryptogram above-mentioned cryptogram of correspondence The characteristic model that history input behavior data are set up, when being input into identical cryptogram due to different user, Corresponding input behavior feature is often the presence of larger difference, and above-mentioned input behavior characteristic model energy Enough reflect that the validated user is input into the rule of the input behavior feature of above-mentioned cryptogram；So, legal After the cryptogram of user is stolen by disabled user, the line of input during password is input due to disabled user It is characterized and there is larger difference with the input behavior feature of validated user, this causes that disabled user's is defeated Enter the rule that behavioural characteristic does not meet the input behavior feature of features described above model reaction, and then cause Above-mentioned disabled user can not by the second certification, therefore, the embodiment of the present invention can improve user authentication Security.

Embodiment of the method one

Reference picture 1, flow chart the step of show a kind of user authen method embodiment one of the invention, Specifically may include steps of：

The input behavior feature of step 101, the cryptogram of collection active user's input；

Step 102, the cryptogram to active user input carry out the first certification；

Step 103, after first certification passes through, according to the corresponding feature of the input behavior feature Model carries out the second certification to the input behavior feature, to obtain corresponding second authentication result, in institute When stating the second certification and passing through, active user's certification success is determined, it is obstructed out-of-date in second certification, Determine active user's authentification failure；Wherein, the characteristic model is according to cryptogram correspondence account The validated user at family is when the cryptogram or cryptogram correspondence cryptogram fragment is input into History input behavior feature set up model.

The embodiment of the present invention goes for the user authentication under any application scenarios.For example, of the invention Embodiment is applied to the user authentication of application program；Wherein, application program can be completed by input method system User authentication, specifically, active user's input cryptogram by after the first certification, it is defeated Entering method system can be according to the defeated of the validated user of the cryptogram corresponding account of input method system storage Enter behavioural characteristic, the input behavior feature of the active user's correspondence cryptogram to being gathered is authenticated, And then the second certification is carried out to active user, because input behavior is characterized in the personalization spy of each user Levy, namely often there is larger difference in the input behavior feature of different user, therefore can be according to upper The input behavior feature for stating active user's correspondence cryptogram carries out the second certification to active user.Can be with Understand, any application that above-mentioned application program can be lodged by input method system, the present invention is real Example is applied not to be any limitation as the specific executive agent of specific application program and user authentication.

The user authentication of the embodiment of the present invention can specifically include：For the first certification rank of cryptogram Second authentication phase of section and the input behavior feature for the corresponding cryptogram of active user；Its In, above-mentioned cryptogram can be user by input units such as dummy keyboard, touch-screen, physical keyboards The conventional cipher of input, it has the characteristic for being easily stolen and intercepting；Above-mentioned input behavior feature can For representing behavioural characteristic during active user's above-mentioned cryptogram of input, for example：Key press time, button Coordinate, key press etc., it is assumed that active user input cryptogram be：Wearefamily, then Corresponding key press time can be：T1, t2, t3, t4, t5, t6, t7, t8, t9, t10, t11；It is right The button coordinate answered can be：P1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11；Corresponding key press can be：N1, n2, n3, n4, n5, n6, n7, n8, n9, N10, n11；The embodiment of the present invention is not specifically limited to the input behavior feature of the cryptogram.

It should be noted that the screen connection mode of above-mentioned cryptogram includes：Click on or upper drawing above is shielded Etc. mode, the embodiment of the present invention is not specifically limited to the screen connection mode of above-mentioned cryptogram；Wherein, In the embodiment of the present invention so that cryptogram in click mode by shielding as an example, the embodiment of the present invention is subject to Explanation；When above-mentioned cryptogram is using upper stroke screen connection mode, the characteristic model of above-mentioned key press and press The characteristic model of key coordinate respectively can using the key press model that is adapted with upper stroke screen connection mode and Button coordinate model, the embodiment of the present invention will not be repeated here.

It should be noted that features described above model can according to the cryptogram corresponding account it is legal History input when cryptogram described in user input or cryptogram correspondence cryptogram fragment The model that behavioural characteristic is set up.Wherein, the input scene of above-mentioned historical behavior feature can specifically include： The scene of user authentication and/or the scene of non-user certification；The scene of above-mentioned user authentication refers to be stepped in user Application program is above-mentioned for the scene that the account and cryptogram of user are authenticated before recording certain application program The scene of non-user certification refers to the input scene in addition to the scene of user authentication, and such as user exists The scene of information input is carried out in OFFICE programs, and for example, user enters row information in browser program The scene of input, for another example, user carries out scene of information input etc. in instant communication program.Can manage Solution, the embodiment of the present invention is not any limitation as the specific input scene of historical behavior feature.

In a kind of alternative embodiment of the invention, the input behavior feature includes：Key press time, press At least one in key coordinate and key press；Wherein,

In the embodiment of the present invention, above-mentioned key press time can specifically characterize user in the button of current button Carving can record often with the time interval between the button moment of adjacent last button, input method system At the button moment of individual button, the time interval between each two adjacent key is then calculated, wherein, on Stating button can specifically include space button；For example：The cryptogram of active user's input is " we are ", Then button moment of each button of input method system record be：139.6,157.9,178.5,199.9,210.1, 222.2；Then the corresponding key press time of above-mentioned cryptogram " we are " can be： 0(139.6-139.6),18.3(157.9-139.6),20.6(178.5-157.9),21.4(199.9-178.5),10.2(210. 1-199.9),12.1(222.2-210.1)；

In another alternative embodiment of the embodiment of the present invention, if including space in above-mentioned cryptogram, Space button can then be extractd, and determine that the key press time of button behind space is the button before space to sky / 2nd of the time at intervals of the button after lattice.As in table 1, the key press time of button " w " is 0； The key press time of button " e " is 18.3；Then the embodiment of the present invention can extract button " Space "；Button The key press time of " a " is 21 ((199.9-157.9)/2)；The key press time of button " r " is 10.2；Button " e " Key press time be 12.1.

Table 1

Button	Moment	Key press time
			w	139.6	0(139.6-139.6)
e	157.9	18.3(157.9-139.6)
			Space	178.5	--
a	199.9	21((199.9-157.9)/2)
			r	210.1	10.2(210.1-199.9)
e	222.2	12.1(222.2-210.1)

In the embodiment of the present invention, above-mentioned button coordinate can specifically represent that active user should in click, touching The particular location of button；Above-mentioned button coordinate is specifically as follows coordinate pos (x_i,y_i), wherein, i can be used In the mark for representing button；In the embodiment of the present invention, specifically coordinate central point can be mapped to each The location sets center of button, obtains being uniformly distributed key zone centered on key position cluster center Domain, each key area corresponding region label pos_id (i), and each key area correspondence one Position distribution coordinate pos (xi, yi)；So, the region labeling that each button can all be obtained Pos_id (i), it is possible to which corresponding position distribution coordinate is searched according to the region labeling, it is corresponding to obtain Button coordinate, for example：Reference picture 2, shows a kind of schematic diagram of key area distribution of the invention, its In, region labeling be 1 region be located at key position cluster center, it can be the origin of coordinates (0,0)； Region labeling is 2 region, and its corresponding coordinate can be (1,0)；Region labeling is 3 region, Its corresponding coordinate can be (- 1,0)；Region labeling is 4 region, and its corresponding coordinate can be (0,1)；Region labeling is 5 region, and its corresponding coordinate can be (1,1)；Region labeling is 6 Region, its corresponding coordinate can be (- 1,1)；Region labeling is 7 region, its corresponding coordinate Can be (0, -1)；Region labeling is 8 region, and its corresponding coordinate can be (1, -1)；Mark in region Number for 9 region, its corresponding coordinate can be (- 1, -1)；Assuming that active user's click keys region Region marked as 4, then can obtain corresponding position distribution coordinate for (0,1)；Assuming that user clicks on The region labeling of button is 5 region, then can obtain corresponding position distribution coordinate for (1,1).

It is appreciated that above-mentioned key area is divided into 1 to 9 totally 9 regions, only as the embodiment of the present invention In the distribution of above-mentioned kind of key area a kind of example, without being not understood as to key zone in the embodiment of the present invention A kind of restriction of domain distribution, in fact, key area can be with Arbitrary distribution, such as：By key area point Into 1 to 4 totally 4 regions；Distribution of the embodiment of the present invention to above-mentioned key area is not specifically limited.

In the embodiment of the present invention, above-mentioned key press can specifically be represented to be clicked on to active user, touches and be somebody's turn to do The key press used during button, can specifically obtain above-mentioned button pressure by pressure sensitive hardware Power；In the embodiment of the present invention, input method system can specifically obtain above-mentioned pressure sensitive hardware and can perceive Minimum pressure and maximum pressure, and determine that the minimum pressure and maximum pressure are pressure reference, to utilize The pressure reference is normalized to the key press of each button, with the button pressure after being normalized Power；For example：Pressure minimum pressure is N1, and maximum pressure is N2, and the pressure of the button of acquisition is N3, Key press after being then normalized to N3 is (N3-N1)/(N2-N1).

It is appreciated that the method for normalizing in above-mentioned example is only as method for normalizing in the embodiment of the present invention A kind of example, without being not understood as in the embodiment of the present invention restriction to method for normalizing, in fact, Normalized method has various, for example：Pressure value is changed into the decimal between (0,1) etc. method, The embodiment of the present invention is not specifically limited to above-mentioned normalized method.

In the embodiment of the present invention, the corresponding characteristic model of above-mentioned input behavior feature can be according to above-mentioned close History input behavior data when the corresponding validated user history of code text is input into above-mentioned cryptogram and build Vertical characteristic model；Because above-mentioned history input behavior data can use input for above-mentioned validated user The input behavior characteristic that method system is accumulated during carrying out above-mentioned cryptogram input, therefore on Stating the corresponding characteristic model of input behavior feature can reflect that above-mentioned validated user is input into above-mentioned cryptogram When input behavior feature rule, and the input behavior feature of different user input identical cryptogram Generally there is larger difference, so, even if knowing the cryptogram of a validated user in disabled user In the case of, also cannot be by imitating the input behavior feature of the validated user and then recognizing by second Card, so the solution of the present invention can effectively ensure that the security of validated user account, prevents illegal use The login at family, so as to improve the security of user authentication.

For example, Li Ming is input into the account and password of Han Mei plums using mobile phone on social network sites, due to Lee The input behavior feature of the bright above-mentioned password of input does not meet the input behavior spy that Han Meimei is input into above-mentioned password Rule is levied, Gu Liming cannot be by the second certification, that is, feelings of the Li Ming in the password for knowing Han Meimei Under condition, also accounts of the Han Meimei on social network sites cannot be logged in by illegal way.

It should be noted that the embodiment of the present invention is not specifically limited to characteristic model, above-mentioned conjunction can be reflected The model of the rule of input behavior feature during the above-mentioned cryptogram of method user input is feasible, example Such as：The models such as gauss hybrid models, Markov chain model, the embodiment of the present invention is mainly with characteristic model To be illustrated as a example by gauss hybrid models, other kinds of characteristic model is cross-referenced.

In addition, it is necessary to explanation, in the embodiment of the present invention, the cryptogram of active user's input can be with It is the corresponding complete password of user account, or the corresponding segmentation password of user account.Assuming that complete Password is " wearefamily ", then can be complete by this after active user disposably input complete password Password verified as cryptogram, or, active user can also be segmented input segmentation password, and The segmentation password is verified as cryptogram, for example, user can divide 3 sections of inputs above-mentioned complete Password, corresponding cryptogram fragment is respectively " we ", " are " and " family ", in such cases completely User authentication process can specifically include corresponding three user authentication process of three cryptogram fragments.

The embodiment of the present invention, above-mentioned cryptogram fragment can be used to represent a part for cryptogram.Specifically Ground, it can be that user carries out the corresponding cryptogram fragment that segmentation input is obtained to cryptogram, Can be that the corresponding cryptogram fragment that cutting is obtained is carried out to the cryptogram being input into, the present invention is implemented Example is not specifically limited to specific cryptogram fragment.

In actual applications, features described above model can specifically include：One according to the cryptogram Input behavior feature and the characteristic model set up or special according to multiple input behavior of the cryptogram The characteristic model levied and set up.

1. the characteristic model that an input behavior feature according to the cryptogram is set up

In the embodiment of the present invention, features described above model is specifically as follows above-mentioned according to the input of above-mentioned validated user The model that one input behavior feature of correspondence during cryptogram is set up, that is, the number of features described above model Amount is corresponding with the quantity of input behavior feature, and each characteristic model is with corresponding input behavior feature It is variable；The input behavior feature is carried out according to the input behavior feature corresponding characteristic model then Second certification, the step of to obtain the second authentication result, can specifically include：

Step A1, the input behavior feature is input into corresponding characteristic model, by the feature Model exports corresponding first confidence level；

Step A2, according to corresponding first confidence level of all input behavior features, obtain active user's Confidence level；

Step A3, the confidence level according to active user, judge whether second certification passes through.

It is above-mentioned according to all input behavior features corresponding first in a kind of alternative embodiment of the invention Confidence level, the step of obtain the confidence level of active user, can specifically include：Calculate the input behavior The average of the corresponding confidence level of feature, and determine the confidence level that the average is active user.

In the embodiment of the present invention, because characteristic model is to use input method system according to above-mentioned validated user Accumulate during above-mentioned cryptogram input input behavior characteristic is set up, it can be with The rule of input behavior feature when reflecting that above-mentioned validated user is input into above-mentioned cryptogram, therefore will be above-mentioned Input behavior feature is input into corresponding characteristic model, namely to current input behavior feature and legal The difference of the accumulative input behavior feature of user's history is considered, the confidence level of features described above model output It is the measurement for weighing above-mentioned difference.

For example：Assuming that input behavior feature includes：Key press time, button coordinate, key press, pass through The key press time confidence level of corresponding characteristic model output is s1, button coordinate confidence level is s2, button pressure Power confidence level is s3；Then according to above-mentioned key press time confidence level be s1, button coordinate confidence level be s2, The step of key press confidence level obtains the confidence level of active user for s3 can specifically include：Will be above-mentioned Key press time confidence level be s1, button coordinate confidence level be s2, key press confidence level be s3 as one Individual three-dimensional vector is input into confidence level Fusion Model, and active user is exported by the confidence level Fusion Model Confidence level；

In the embodiment of the present invention, above-mentioned confidence level Fusion Model is specifically as follows model of averaging, then melt Close key press time confidence level be s1, button coordinate confidence level be s2, key press confidence level be s3 with To the confidence level of active user, it is specifically as follows：S=(s1+s2+s3)/3；Wherein, above-mentioned s is specific The confidence level of active user can be represented.

It is appreciated that obtaining the confidence level of active user only as the present invention above by model of averaging According to corresponding first confidence level of all input behavior features in embodiment, the confidence level of active user is obtained Mode a kind of example, without be not understood as in the embodiment of the present invention according to all input behavior features Corresponding first confidence level, obtains a kind of limitation of the mode of the confidence level of active user, in fact, also Can pass through：Weighted average, different characteristic set different weights, complicated Multilayer Perception network or The methods such as person's SVMs obtain the confidence level of active user, and the embodiment of the present invention is to according to all defeated Enter corresponding first confidence level of behavioural characteristic, the mode for obtaining the confidence level of active user does not make specific limit It is fixed.

It should be noted that above-mentioned Fusion Model specifically needs to carry out Training process, specifically there is prison Training process is superintended and directed with reference to prior art, the embodiment of the present invention is repeated no more to this.

Using example one

In the embodiment of the present invention, the input behavior feature can specifically include：Key press time, and/or press Key coordinate, and/or key press；The characteristic model can specifically include：Key press time model and/ Or button coordinate model, and/or key press model；Then according to the corresponding spy of the input behavior feature Levy model carries out the second certification to the input behavior feature, to obtain corresponding second authentication result Step, can specifically include：

The key press time is input into the key press time model, is exported by the key press time model The key press time confidence level of active user；And/or

The button coordinate is input into the button coordinate model, is exported by the button coordinate model The button coordinate confidence level of active user；And/or

The key press is input into the key press model, is exported by the key press model The key press confidence level of active user；

Sat according to the corresponding key press time confidence level of all input behavior features, and/or the button Mark confidence level, and/or the key press confidence level, obtain the confidence level of active user；

For example：Cryptogram is " weare ", and the input behavior feature of cryptogram includes：Key press time, Button coordinate and key press, respectively：Key press time " t1, t2, t3, t4, t5 "；Button coordinate " p1, p2, p3, p4, p5 "；Key press " n1, n2, n3, n4, n5 "；Then pressed above-mentioned respectively Key time, button coordinate, key press are input into corresponding key press time characteristic model as variable, press In key coordinate model, key press model, put with obtaining corresponding key press time confidence level, button coordinate Reliability, button coordinate confidence level, and according to the corresponding key press time confidence level of above-mentioned input behavior feature, Button coordinate confidence level, button coordinate confidence level, obtain the confidence level of active user.

2. the characteristic model that the multiple input behavior features according to the cryptogram are set up

In the embodiment of the present invention, features described above model is specifically as follows above-mentioned according to the input of above-mentioned validated user The model that corresponding multiple input behavior features are set up during cryptogram, therefore, the quantity of characteristic model can Think one, namely characteristic model is characterized as variable with multiple input behavior of cryptogram；Then according to institute State the corresponding characteristic model of input behavior feature carries out the second certification to the input behavior feature, with The step of to the second authentication result, can specifically include：

Step D1, the multiple input behavior feature is input into corresponding characteristic model, by described Characteristic model exports the confidence level of active user；

Step D2, the confidence level according to active user, judge whether second certification passes through.

Using example two

In the embodiment of the present invention, the input behavior feature includes：Key press time, and/or button coordinate, And/or key press；Then it is described according to the corresponding characteristic model of the input behavior feature to the input Behavioural characteristic carries out the second certification, the step of to obtain corresponding second authentication result, can specifically wrap Include：The key press time, and/or button coordinate, and/or key press are input into the characteristic model In, the confidence level of the active user is exported by the characteristic model.

For example：Cryptogram is " weare ", and the input behavior feature of cryptogram includes：Key press time, Button coordinate and key press, respectively：Key press time " t1, t2, t3, t4, t5 "；Button coordinate " p1, p2, p3, p4, p5 "；Key press " n1, n2, n3, n4, n5 "；When then by above-mentioned button Between, button coordinate, key press be input into a characteristic model as variable, the output result for obtaining The as confidence level of active user.

The present invention can provide the following manner for setting up characteristic model：

The average and variance of history input behavior feature or history input behavior characteristic fragment are calculated, to build The vertical gauss hybrid models with the average and variance as parameter.

Mode one

The process that mode one sets up characteristic model can specifically include：Collect the corresponding history of cryptogram defeated Enter behavioural characteristic, the average and variance of above-mentioned history input behavior feature are calculated, to set up with the average It is the gauss hybrid models of parameter with variance.

Assuming that the input behavior feature in the embodiment of the present invention can specifically include：Key press time, button are sat Mark and key press, then data that can respectively according to history key press time, the data of history button coordinate, The data of history key press calculate the average and variance of correspondence input behavior feature, namely according to history The data of key press time calculate the average one and variance one of key press time, to set up with above-mentioned average one and side Differ from a fisrt feature model for being parameter；Namely calculate button coordinate according to the data of history button coordinate Average two and variance two, to set up the second feature model with above-mentioned average two and variance two as parameter； Namely according to the data of history key press calculate key press average three and variance three, with set up with Above-mentioned average three and variance three are the third feature model of parameter；

Assuming that cryptogram is " weare ", user is input into the cryptogram in register account number, adopts twice Key press time 1 when the correspondence for integrating is input into cryptogram " weare " for the first time is t1, t2, t3, t4, t5；Key press time 2 when the correspondence of collection is input into cryptogram " weare " for the second time is t6, t7, T8, t9, t10；Above-mentioned button is then calculated according to above-mentioned key press time 1 and key press time 2 respectively Average (the t of " w " " e " " a " " r " " e " corresponding key press time_w, t_e1, t_a, t_r, t_e2) and variance (vt_w, vt_e1, vt_a, vt_r, vt_e2), to set up with above-mentioned average (t_w, t_e1, t_a, t_r, t_e2) and variance (vt_w, vt_e1, vt_a, vt_r, vt_e2) it is the fisrt feature model of parameter；For example：The average of the key press time of above-mentioned button " w " can Think t_w=(t1+t6)/2；Variance can be vt_w=((t1-t_w)*(t1-t_w)+(t6-t_w)*(t6-t_w))/2； The average of the key press time of above-mentioned button " e " can be t_e1=(t2+t7)/2；Variance can be vt_e1=((t2-t_e1)*(t2-t_e1)+(t7-t_e1)*(t7-t_e1))/2；

It is appreciated that above-mentioned set up characteristic model only as this hair according to key press time 1 and key press time 2 A kind of example for setting up characteristic model in bright embodiment, without being not understood as building the embodiment of the present invention A kind of restriction of vertical characteristic model, indeed, it is possible to the multiple history produced according to multiple input behavior Button behavioural characteristic data set up the characteristic model, and the embodiment of the present invention do not do specific limit to this It is fixed.

The active user of collection to should cryptogram button coordinate：Correspondence input the password for the first time Button coordinate 1 during text " weare " is：P1, p2, p3, p4, p5；Second input of correspondence should Button coordinate 2 during cryptogram " weare " is p6, p7, p8, p9, p10；

Above-mentioned button is then calculated according to above-mentioned button coordinate 1 and button coordinate 2 respectively Average (the p of " w " " e " " a " " r " " e " corresponding button coordinate_w, p_e1, p_a, p_r, p_e2) and variance (vp_w, vp_e1, vp_a, vp_r, vp_e2), to set up with above-mentioned average (p_w, p_e1, p_a, p_r, p_e2) and variance (vp_w, vp_e1, vp_a, vp_r, vp_e2) it is the second feature model of parameter；

The active user of collection to should cryptogram key press：Correspondence input the password for the first time Key press 1 during text " weare " is：N1, n2, n3, n4, n5；Second input of correspondence should Key press 2 during cryptogram " weare " is n6, n7, n8, n9, n10；Then according to above-mentioned button Pressure 1 and key press 2 calculate the corresponding key press of above-mentioned button " w " " e " " a " " r " " e " respectively Average (n_w, n_e1, n_a, n_r, n_e2) and variance (vn_w, vn_e1, vn_a, vn_r, vn_e2), to set up With above-mentioned average (n_w, n_e1, n_a, n_r, n_e2) and variance (vn_w, vn_e1, vn_a, vn_r, vn_e2) be The third feature model of parameter.

Mode two

The process that mode two sets up characteristic model can specifically include：Collect that cryptogram fragment is corresponding goes through History input behavior characteristic fragment, calculate history input behavior characteristic fragment average and variance, with set up with The average and variance are the gauss hybrid models of parameter.

For length is more long or sentence on for the poor cryptogram of continuity, it recognizes in non-user The probability being transfused under the scene of card is relatively small, for example：Cryptogram is " wearefamily ", legal The number of times that user may be input into " wearefamily " in daily behavior is less, therefore corresponding history is input into Behavioral data is less, is so easily caused the situation of history input behavior Sparse, and the Sparse Easily causing cannot set up characteristic model corresponding with cryptogram, or the differentiation rate of characteristic model is relatively low The problems such as.

In aforesaid way two, above-mentioned cryptogram can be cut into the shorter cryptogram piece of multiple length Section, to increase the probability that cryptogram fragment is transfused to, and is conducive to accumulative above-mentioned password text fragment correspondence History input behavior data, be thus advantageous to set up corresponding characteristic model and characteristic model can be increased Differentiation rate.For example：By above-mentioned " wearefamily " be cut into binary cryptogram fragment " we ", " ea ", After " ar ", " re ", " rf ", " fa ", " am ", " mi ", " il ", " ly ", it is with cryptogram fragment " we " Example, when user input " wecan ", " wehave " etc. text, can collect cryptogram fragment " we " corresponding input behavior data, therefore it is easy to input method system to add up enough cryptogram fragments History input behavior data, differentiation rate input behavior characteristic fragment model higher is set up, that can improve The degree of accuracy of the second certification is carried out to active user.

Can be equal length by above-mentioned cryptogram cutting in a kind of alternative embodiment of the invention Cryptogram fragment, then step B1 can specifically include：

Step C1, the cryptogram is cut into N units cryptogram fragment, by the cryptogram History input behavior feature be cut into N units history input behavior characteristic fragment；Wherein, N be more than Integer equal to 1, the length for representing cryptogram fragment and history input behavior characteristic fragment；

Step C2, the average and variance that calculate the history input behavior characteristic fragment, to set up with institute It is the characteristic model of parameter to state average and variance.

Assuming that above-mentioned cryptogram is " weare " in the embodiment of the present invention, above-mentioned N is 2, namely cutting Cryptogram fragment length be 2, then above-mentioned " weare " is cut into 2 yuan of cryptogram fragments For：" we ", " ea ", " ar ", " re "；User is input into the cryptogram in register account number twice, collection Key press time 1 of correspondence when being input into cryptogram " weare " for the first time be t1, t2, t3, t4, t5；Key press time 2 when the correspondence of collection is input into cryptogram " weare " for the second time is t6, t7, T8, t9, t10；Above-mentioned key press time is then cut into 2 yuan of input behavior characteristic fragments：Key press time 1 2 yuan of input behavior characteristic fragments be " t1, t2 ", " t2, t3 ", " t3, t4 ", " t4, t5 "；Button 2 yuan of input behavior characteristic fragments of time 2 are " t6, t7 ", " t7, t8 ", " t8, t9 ", " t9, t10 "；

Then according to above-mentioned " we " corresponding key press time 1 and 2 yuan of input behavior feature pieces of key press time 2 Section：" t1, t2 " and " t6, t7 " calculates the equal of the corresponding key press time of above-mentioned button " w " " e " respectively Value (t_w, t_e1) and variance (vt_w, vt_e1), to set up with above-mentioned average (t_w, t_e1) and variance (vt_w, vt_e1) it is the fisrt feature model 1 of parameter；

Then according to above-mentioned " ea " corresponding key press time 1 and 2 yuan of input behavior feature pieces of key press time 2 Section：" t2, t3 " and " t7, t8 " calculates the average of the corresponding key press time of above-mentioned button " e " " a " respectively (t_e1, t_a) and variance (vt_e1, vt_a), to set up with above-mentioned average (t_e1, t_a) and variance (vt_e1, vt_a) it is the fisrt feature model 2 of parameter；

Then according to above-mentioned " ar " corresponding key press time 1 and 2 yuan of input behavior feature pieces of key press time 2 Section：" t3, t4 " and " t8, t9 " calculates the average of the corresponding key press time of above-mentioned button " a " " r " respectively (t_a, t_r) and variance (vt_a, vt_r), to set up with above-mentioned average (t_a, t_r) and variance (vt_a, vt_r) It is the fisrt feature model 3 of parameter；

Then according to above-mentioned " re " corresponding key press time 1 and 2 yuan of input behavior feature pieces of key press time 2 Section：" t4, t5 " and " t9, t10 " calculates the equal of the corresponding key press time of above-mentioned button " r " " e " respectively Value (t_r, t_e2) and variance (vt_r, vt_e2), to set up with above-mentioned average (t_r, t_e2) and variance (vt_r, vt_e2) it is the fisrt feature model 4 of parameter；

Process is set up for button coordinate and key press character pair model, during due to it with button Between character pair model set up process be similar to, and therefore not to repeat here, cross-referenced.

Can be that length is unequal by above-mentioned cryptogram cutting in a kind of alternative embodiment of the invention Cryptogram fragment, then step B1 can specifically include：

The history input behavior feature of the cryptogram and the cryptogram is carried out into M grades to cut Point, the cryptogram is cut into first password text fragments, the second cryptogram fragment, M Cryptogram fragment；Wherein, the first password text fragments are N units cryptogram fragment；It is described Second cryptogram fragment is O units cryptogram fragment；The M cryptograms fragment is that P units are close Code text fragments；Wherein, above-mentioned N is the integer more than or equal to 1；O is the integer more than N；P It is the integer more than O.

In the embodiment of the present invention, it is assumed that cryptogram is " weare ", user is defeated twice in register account number Enter the cryptogram, the correspondence of collection is input into the key press time 1 during cryptogram " weare " for the first time It is t1, t2, t3, t4, t5；The correspondence of collection is input into button during cryptogram " weare " for the second time Time 2 is t6, t7, t8, t9, t10；M is that 3, N is that 2, O is that 3, P is 4, will also be gone up Stating cryptogram and key press time carries out 3 grades of cuttings, and the password of binary, ternary and quaternary is cut into respectively It is " we ", " wea ", " eare " after text fragments, wherein cryptogram " weare " cutting；Key press time 1 It is after cutting：" t1, t2 ", " t1, t2, t3 ", " t2, t3, t4, t5 "；It is after the cutting of key press time 2： " t6, t7 ", " t6, t7, t8 ", " t7, t8, t9, t10 "；

Then according to above-mentioned " wea " corresponding key press time 1 and 2 yuan of input behavior features of key press time 2 Fragment：" t1, t2, t3 " and " t6, t7, t8 " calculates corresponding the pressing of above-mentioned button " w " " e " " a " respectively Average (the t of key time_w, t_e1, t_a) and variance (vt_w, vt_e1, vt_a), to set up with above-mentioned average (t_w, t_e1, t_a) and variance (vt_w, vt_e1, vt_a) it is the fisrt feature model 2 of parameter；

Then according to above-mentioned " eare " corresponding key press time 1 and 2 yuan of input behavior features of key press time 2 Fragment：" t2, t3, t4, t5 " and " t7, t8, t9, t10 " calculates above-mentioned button respectively Average (the t of the corresponding key press time of " e " " a " " r " " e "_e1, t_a, t_r, t_e2) and variance (vt_e1, vt_a, vt_r, vt_e2), to set up with above-mentioned average (t_e1, t_a, t_r, t_e2) and variance (vt_e1, vt_a, vt_r, vt_e2) It is the fisrt feature model 3 of parameter；

Foundation of the characteristic model of button coordinate and key press with reference to the characteristic model of above-mentioned key press time Process.

In a kind of alternative embodiment of the invention, step can also be included：

Set up the rope with the cryptogram as major key, with the storage location of the characteristic model as storage value Draw.

Reference picture 3, show it is provided in an embodiment of the present invention it is a kind of set up characteristic model method the step of Flow chart, can specifically include：

Step 301, input method system receive the cryptogram of validated user input；

Step 302, input method system record the cryptogram of validated user input, key press time, Button coordinate and key press；

Step 303, input method system according to the corresponding key press time of the cryptogram, button coordinate, Key press sets up corresponding key press time model, button coordinate model and key press model；

Wherein, above-mentioned steps 303 can specifically include step：

The index of step 3031, foundation with cryptogram as major key；

Wherein, the step of index of the foundation with cryptogram as major key, can specifically include：

Step T1, the cryptogram is cut into N units cryptogram fragment, to obtain multiple N units Cryptogram fragment；

Step T2, corresponding index is set up according to the cryptogram fragment；

Step 3032, the key press time model is set up according to the key press time；

Wherein, it is described the step of set up the key press time model according to the key press time, specifically can be with Including：

Step Y1, the normalization key press time, and when the key press time is cut into N units button Between fragment, to obtain multiple N unit key press time fragment；

Step Y2, set up between N unit's cryptogram fragment and N units key press time fragment Corresponding relation；

Step Y3, corresponding N unit key press time fragment mould is set up according to N units key press time fragment Type；

Step Y4, foundation are with N units cryptogram fragment as major key, with N units cryptogram fragment The storage location of corresponding N units key press time piece segment model is the index of storage value；

Step 3033, the button coordinate model is set up according to the button coordinate；

Wherein, it is above-mentioned the step of set up the button coordinate model according to the button coordinate, specifically can be with Including：

Step X1, the normalization button coordinate, and the button coordinate is cut into N units button seat Tap section, to obtain multiple N units button coordinate fragment；

Step X2, set up between N unit's cryptogram fragment and N units button coordinate fragment Corresponding relation；

Step X3, corresponding N unit button coordinate fragment mould is set up according to N units button coordinate fragment Type；

Step X4, foundation are with N units cryptogram fragment as major key, with N units cryptogram fragment The storage location of corresponding N units button coordinate piece segment model is the index of storage value；

Step 3034, the key press model is set up according to the key press；

Wherein, it is above-mentioned the step of set up the key press model according to the key press, specifically can be with Including：

Step E1, the normalization key press, and the key press is cut into N units button pressure Power fragment, to obtain multiple N units key press fragment；

Step E2, set up between N unit's cryptogram fragment and N units key press fragment Corresponding relation；

Step E3, corresponding N unit key press fragment mould is set up according to N units key press fragment Type；

Step E4, foundation are with N units cryptogram fragment as major key, with N units cryptogram fragment The storage location of corresponding N units key press piece segment model is the index of storage value.

It should be noted that in the embodiment of the present invention, above-mentioned cryptogram is " weare " only as the present invention A kind of example of cryptogram in embodiment, without being not understood as to the cryptogram in the embodiment of the present invention A kind of restriction, in fact, in cryptogram in addition to letter, numeral, punctuate and symbol can also be included Number, the embodiment of the present invention is not specifically limited to cryptogram.

In the embodiment of the present invention, when the result of above-mentioned second certification passes through for the second certification, then can be with Determine the authentication success of active user, otherwise determine the authentication failure of active user.

To sum up, the embodiment of the present invention can be defeated when the input of cryptogram is carried out according to active user Enter behavioural characteristic to carry out user authentication to user, be specifically as follows and the first certification is passed through to cryptogram User, the second certification is carried out according to the corresponding characteristic model of cryptogram；Features described above model specifically may be used Being the history input behavior data according to the validated user of the above-mentioned cryptogram above-mentioned cryptogram of correspondence The characteristic model of foundation, when being input into identical cryptogram due to different user, corresponding input behavior Feature is often the presence of larger difference, and above-mentioned input behavior characteristic model can reflect the legal use Family is input into the rule of the input behavior feature of above-mentioned cryptogram；So, in the cryptogram of validated user After being stolen by disabled user, input behavior feature and legal use during the password are input due to disabled user There is larger difference in the input behavior feature at family, this causes that the input behavior feature of disabled user is not inconsistent Close and state the rule of the input behavior feature that characteristic model is reacted, and then cause above-mentioned disabled user not The second certification can be passed through, therefore, the embodiment of the present invention can improve the security of user authentication.

Embodiment of the method two

Reference picture 4, flow chart the step of show a kind of user authen method embodiment two of the invention, Specifically may include steps of：

The input behavior feature of step 401, the cryptogram of collection active user's input；

Step 402, the cryptogram to active user input carry out the first certification；

Step 403, after first certification passes through, search index with the cryptogram as major key； Wherein, the index, with the cryptogram as major key, is storage with the storage location of the characteristic model Value；

Step 404, the storage location of the characteristic model is obtained according to described index, and obtain corresponding Characteristic model；

Step 405, according to the corresponding characteristic model of the input behavior feature to the input behavior feature The second certification is carried out, to obtain corresponding second authentication result, when second certification passes through, really Settled preceding user authentication success, it is obstructed out-of-date in second certification, determine that active user's certification is lost Lose；Wherein, the characteristic model is the validated user according to the cryptogram corresponding account in input History input behavior feature when the cryptogram or cryptogram correspondence cryptogram fragment The model of foundation.

Relative to embodiment of the method one, the embodiment of the present invention increased step 403 and step 404, the step Rapid 403 and step 404 in can search index with the cryptogram as major key；Due to can basis The index quickly obtains corresponding characteristic model, therefore can improve the speed for obtaining characteristic model, so that The speed of the second certification can be improved.

In the embodiment of the present invention, if being found with cryptogram in database according to above-mentioned cryptogram Be the index of major key, then can from above-mentioned index the corresponding characteristic model of cryptogram storage location Obtain the characteristic model；

In the embodiment of the present invention, if features described above model is the multiple input behavior according to the cryptogram Feature set up characteristic model, then the index with above-mentioned cryptogram as major key can correspond to a character modules The storage location of type, namely according to the storage location for indexing and obtaining the characteristic model, and then obtain The characteristic model quantity can be one；If features described above model is according to the cryptogram The characteristic model that one input behavior feature is set up, then the index with above-mentioned cryptogram as major key can be right The storage location of multiple characteristic models is answered, namely according to the storage position for indexing and obtaining the characteristic model Put, and then the quantity of the characteristic model for obtaining can be multiple.

In a kind of alternative embodiment of the invention, the above method also includes step：

Step P1, when the index with the cryptogram as major key is not found, search with described close The depression of order text of code text is the first index of major key；

Step P2, the storage location of characteristic model is obtained according to the described first index, and obtain corresponding Characteristic model.

In the embodiment of the present invention, the depression of order text of above-mentioned cryptogram is the password text of above-mentioned cryptogram This fragment；Said process is illustrated below by way of specific example：Assuming that cryptogram is： " weare ", when manipulative indexing is not found according to " weare ", can carry out depression of order to " weare " Treatment, to obtain the depression of order text of " weare ", for example：Can by " weare " depression of order turn into " we ", " ea ", " ar ", " re ", and looking into for the first index carried out according to above-mentioned " we ", " ea ", " ar ", " re " respectively Look for；Because the string length of " weare " is more long, it is thus possible to history input behavior Sparse occurs Situation, cause characteristic model not corresponding with " weare "；" but we ", " ea ", " ar ", " re " String length Deng substring is shorter, and its corresponding history input behavior data will be more accordingly Some, are conducive to setting up corresponding characteristic model, therefore the depression of order text that can be treated according to depression of order Corresponding first index is searched, and then obtains corresponding characteristic model.

Embodiment of the method three

Reference picture 5, flow chart the step of show a kind of user authen method embodiment three of the invention, Specifically may include steps of：

The input behavior feature of step 501, the cryptogram of collection active user's input；

Step 502, the cryptogram to active user input carry out the first certification；

Step 503, after first certification passes through, the cryptogram is cut into multiple passwords text This fragment；Wherein, the cryptogram fragment has corresponding input behavior characteristic fragment；

Step 504, the corresponding input behavior characteristic fragment of the cryptogram fragment is input into corresponding In characteristic model, corresponding second confidence level is exported by the characteristic model；

Step 505, according to corresponding second confidence level of all cryptogram fragments, obtain active user's Confidence level；

Step 506, the confidence level according to active user, judge whether second certification passes through.

Relative to embodiment of the method one, the embodiment of the present invention increased step 503, with to password text This and input behavior feature are cut into multiple cryptogram fragments and input behavior characteristic fragment；And lead to Step 504, step 505 and step 506 are crossed pair according to the corresponding characteristic model of the input behavior feature Second certification is carried out to the input behavior feature, is entered the step of to obtain corresponding second authentication result Refinement is gone.

Because the length of cryptogram is long, (such as cryptogram is character string " wearefamily ", character The character that includes of string is more) when, during the input that user carries out text using input method system daily, tire out The history input behavior data of the above-mentioned cryptogram of meter are less, according to a small amount of history input behavior The degree of accuracy of characteristic model that feature is set up namely relatively low；Therefore above-mentioned cryptogram can be cut Point, also will above-mentioned cryptogram be cut into multiple cryptogram fragments, for example：Cryptogram is： " weare ", 2 yuan of cuttings are carried out by above-mentioned cryptogram, then can be cut into cryptogram " we " " ea " " ar " " re ", because the length of the cryptogram after cutting is shorter, therefore compares beneficial to accumulative Corresponding history input behavior data, accumulative history input behavior data are more, the spy for hence setting up Levy the degree of accuracy of model namely higher.

In a kind of alternative embodiment of the invention, the length of above-mentioned each cryptogram fragment is identical, or Person, the length of each cryptogram fragment is different；

In the embodiment of the present invention, it is assumed that above-mentioned cryptogram " weare ", in each cryptogram fragment Under length identical scene, it is assumed that above-mentioned " weare " is cut into 2 yuan of cryptogram fragments, specially： “we”“ea”“ar”“re”；The active user of collection to should the key press time of cryptogram be：T1, t2, T3, t4, t5；Above-mentioned key press time then is cut into 2 yuan of input behavior characteristic fragments is：" t1, t2 ", " t2, t3 ", " t3, t4 ", " t4, t5 "；According to above-mentioned 2 yuan of cryptogram fragments " we ", " ea ", " ar ", " re " finds corresponding key press time model 1, key press time model 2, key press time model 3 and button , then above-mentioned " t1, t2 " as variable be input into key press time characteristic model 1 by time model 4, with Obtain the time confidence level 1 of cryptogram fragment " we "；When " t2, t3 " is input into button as variable Between in characteristic model 2, to obtain the time confidence level 2 of cryptogram fragment " ea "；" t3, t4 " is made For variable is input into key press time characteristic model 3, to obtain the time confidence of cryptogram fragment " ar " Degree 3；" t4, t5 " is input into key press time characteristic model 4 as variable, to obtain cryptogram The time confidence level 4 of fragment " re "；And it is current to merge above-mentioned time confidence level 1 to time confidence level 4 The confidence level of user.

Under the different scene of the length of each cryptogram fragment, it is assumed that above-mentioned " weare " is cut into 2 yuan, 3 yuan, 4 yuan of cryptogram fragments, specially：“we”、“wea”、“eare”；What is gathered works as Preceding user to should the key press time of cryptogram be：T1, t2, t3, t4, t5；Then by above-mentioned button Time be cut into 2 yuan, 3 yuan, 4 yuan of input behavior characteristic fragments be：" t1, t2 ", " t1, t2, t3 ", " t2, t3, t4, t5 "；Then according to the corresponding characteristic model of the input behavior characteristic fragment to described defeated Entering behavioural characteristic fragment carries out the second certification, to obtain the process of corresponding second authentication result with reference to close Process under the length identical scene of code text fragments.

In the embodiment of the present invention, after above-mentioned cryptogram is cut into multiple cryptogram fragments, by defeated Enter behavioural characteristic fragment model to be authenticated above-mentioned cryptogram fragment, and then realization is entered to active user The certification of row second；Because the length of cryptogram fragment is shorter, therefore can be in the daily line of input of user The history input behavior data of cryptogram fragment are got in, for example：Cryptogram is " wearefamily ", the number of times that validated user is input into " wearefamily " in daily behavior is less, therefore Corresponding history input behavior data are less；But above-mentioned " wearefamily " is cut into binary password text After this fragment " we ", " ea ", " ar ", " re ", " rf ", " fa ", " am ", " mi ", " il ", " ly ", with As a example by cryptogram fragment " we ", when user input " wecan ", " wehave " etc. text, can be with The corresponding input behavior data of cryptogram fragment " we " are collected, therefore is easy to the accumulative foot of input method system The history input behavior data of enough cryptogram fragments, set up degree of accuracy input behavior feature piece higher Segment model, the degree of accuracy of the second certification is carried out that can improve to active user.

Embodiment of the method four

Reference picture 6, flow chart the step of show a kind of user authen method example IV of the invention, Specifically may include steps of：

The input behavior feature of step 601, the cryptogram of collection active user's input；

Step 602, the cryptogram to active user input carry out the first certification；

Step 603, after first certification passes through, the cryptogram is cut into multiple passwords text This fragment；Wherein, the cryptogram fragment has corresponding input behavior characteristic fragment；

The index of step 604, lookup with the cryptogram fragment as major key；Wherein, it is described index with The cryptogram fragment is major key, is storage with the storage location of the input behavior characteristic fragment model Value；

Step 605, the corresponding input behavior characteristic fragment of the cryptogram fragment is input into corresponding In input behavior characteristic fragment model, put by input behavior characteristic fragment model output corresponding second Reliability；

Step 606, according to corresponding second confidence level of all cryptogram fragments, obtain active user's Confidence level；

Step 607, the confidence level according to active user, judge whether second certification passes through.

In the embodiment of the present invention, it is assumed that cryptogram is " weare ", then by above-mentioned cryptogram password text This " weare " is cut into 2 yuan of cryptogram fragments and is：“we”、“ea”、“ar”、“re”；What is gathered works as Preceding user to should cryptogram key press time be t11, t12, t13, t14, t15；Then pressed above-mentioned The key time is cut into 2 yuan of input behavior characteristic fragments：" t11, t12 ", " t12, t13 ", " t13, t14 ", " t14, t15 "；Cutting of the dicing process of button coordinate and key press with reference to above-mentioned key press time Journey.

In the embodiment of the present invention, respectively according to above-mentioned cryptogram fragment：“we”、“ea”、“ar”、“re” It is the index of major key to search with " we ", " ea ", " ar ", " re ", according to the input behavior that above-mentioned index is obtained Piece segment model includes：Temporal characteristics model：Temporal characteristics model 1, temporal characteristics model 2, time are special Levy model 3 and time characteristic model 4；Translation specifications model, including：Translation specifications model 1, seat Mark characteristic model 2, translation specifications model 3 and translation specifications model 4；Pressure characteristic model, bag Include：Pressure characteristic model 1, pressure characteristic model 2, pressure characteristic model 3 and pressure characteristic model 4；

In the embodiment of the present invention, it is above-mentioned get corresponding input behavior characteristic model after, can will be above-mentioned 2 yuan of input behavior characteristic fragments are separately input into corresponding input behavior piece segment model, with temporal characteristics It is illustrated as a example by model：Above-mentioned " t11, t12 " is input into above-mentioned temporal characteristics model 1 as variable In, to obtain the key press time confidence level 1 of active user；Above-mentioned " t12, t13 " is input into as variable Into temporal characteristics model 2, to obtain the key press time confidence level 2 of active user；By it is above-mentioned " t13, T14 " is input into temporal characteristics model 3 as variable, to obtain the key press time confidence level of active user 3；Above-mentioned " t14, t15 " is input into temporal characteristics model 4 as variable, to obtain active user's Key press time confidence level 4；

After above-mentioned 4 key press time confidence levels for obtaining active user, can be by above-mentioned key press time confidence Degree 1 to key press time confidence level 4 is merged, to obtain the key press time confidence level of active user；This The calculating process of above-mentioned button coordinate confidence level and key press confidence level is with reference to above-mentioned in inventive embodiments The calculating process of key press time confidence level；

In the embodiment of the present invention, key press time confidence level, button coordinate confidence level, key press are being obtained After confidence level, can be according to above-mentioned key press time confidence level, button coordinate confidence level, key press confidence Degree, obtains the confidence level of active user；According to above-mentioned key press time confidence level, button coordinate confidence Degree, key press confidence level, obtaining the method for the confidence level of active user has various：For example it is averaging Value, or the methods such as complicated Multilayer Perception network or SVMs, the present invention can be utilized Embodiment is not especially limited to this.

In the embodiment of the present invention, can be carried out to active user according to the confidence level of the active user for obtaining Two certifications, if confidence level is more than confidence threshold value, active user is by the second certification, namely works as Preceding authenticating user identification success；If confidence level is less than or equal to confidence threshold value, active user is not led to Cross the second certification, namely current user identities authentification failure.

In a kind of alternative embodiment of the invention, above-mentioned confidence threshold value is specifically as follows input method system The accumulation of the history confidence level according to the validated user to the cryptogram corresponding account, and then obtain A dynamic threshold；It is specifically as follows, the validated user for accumulating the cryptogram corresponding account is gone through History confidence level, the drafting of distribution map is carried out according to the history confidence level, then by confidence level distribution number compared with Less and the relatively low region of fraction is set to confidence threshold value；For example：The number of the history confidence level of accumulation It is 10, respectively：8、8、8、7、6、9、7、2、6、4；The then scope 0 of corresponding confidence level Into 4, two history confidence levels are only included, therefore confidence threshold value can be set to 4；

In another alternative embodiment of the invention, can be corresponding according to different input behavior features Different weights dynamically set above-mentioned confidence threshold value；For example：Assuming that above-mentioned input behavior feature includes：Press Key time, button coordinate and key press, and 40%, 30%, 30% is accounted in confidence threshold value respectively Weights, if the key press time confidence level of active user be 8, button coordinate confidence level be 7, button pressure Power confidence level is 7, then can set the corresponding confidence threshold value of active user is 8*40%+7*30%+7*30%=7.4；Above-mentioned key press time confidence is merged according to the mode averaged Degree, button coordinate confidence level and key press confidence level be the confidence level of active user, then active user Confidence level is：(8+7+7)/3=7.3, the confidence level of active user is less than confidence threshold value, then current to use The authentification failure of family second.

Reference picture 7, show the present invention offer a kind of user authen method example the step of flow chart, Can specifically include：

Step 701, input method system receive the cryptogram of active user's input, and gather active user The input behavior feature of the cryptogram of input；

Wherein, the step of above-mentioned input method system collection cryptogram corresponding input behavior feature, specifically Can include：Cryptogram, key press time, button coordinate, the key press of record active user's input；

Step 702, input method system carry out the first certification to the cryptogram, with certification current password The correctness of text；If the authentification failure of current password text first, active user's authentification failure；

If step 703, active user are by the first certification, input method system to above-mentioned cryptogram, press Key time, button coordinate, key press carry out background process；

Wherein, above-mentioned input method system is to above-mentioned cryptogram, key press time, button coordinate, button pressure The step of power carries out background process, can specifically include：

Step Q1, the key press time, button coordinate, key press are normalized, and by the button Time, button coordinate, key press be cut into N units key press times fragment, N units button coordinates fragment, N units key press fragment, to obtain multiple N units key press time fragments, the first button coordinate fragments of N, N First key press fragment；

Step Q2, set up between N unit's cryptogram fragment and N units key press time fragment Corresponding pass between corresponding relation, N unit's cryptogram fragments and the first button coordinate fragments of the N Corresponding relation between system, N unit's cryptogram fragments and N units key press fragment；

Step 704, input method system using storage characteristic model to above-mentioned key press time, button coordinate, Key press carries out the second certification；

Step 705, judge the active user whether by the second certification；If the active user second Certification success, then active user's certification success；If the authentification failure of the active user second, institute State active user's authentification failure.

Embodiment of the method five

Reference picture 8, flow chart the step of show a kind of user authen method embodiment five of the invention, Specifically may include steps of：

The input behavior feature of step 801, the cryptogram of collection active user's input；

Step 802, the cryptogram to active user input carry out the first certification；

Step 803, after first certification passes through, according to the corresponding feature of the input behavior feature Model carries out the second certification to the input behavior feature, to obtain corresponding second authentication result, When second certification passes through, active user's certification success is determined, do not pass through in second certification When, determine active user's authentification failure；Wherein, the characteristic model is according to the cryptogram pair The validated user of account is answered to be input into the cryptogram or cryptogram correspondence cryptogram piece The model that history input behavior feature during section is set up；

Step 804, when second certification passes through, according to active user correspondence password text This input behavior feature carries out the first renewal to the corresponding characteristic model of the input behavior feature, or Person, the input behavior characteristic fragment according to cryptogram fragment is corresponding to the input behavior characteristic fragment Characteristic model carries out the second renewal.

Relative to embodiment of the method one, the embodiment of the present invention increased step 804, can be according to the step The corresponding characteristic model of the 804 pairs of input behavior features carries out the first renewal, or to input behavior feature piece The corresponding characteristic model of section carries out the second renewal, to improve the degree of accuracy of characteristic model.

In the embodiment of the present invention, features described above model is specifically as follows gauss hybrid models, due to the height The parameter of this mixed model is the average and variance of history input behavior data, therefore above-mentioned Gauss is mixed The renewal of matched moulds type namely the renewal to above-mentioned average and variance, then the step of the first renewal, specifically may be used To include：

The step of second renewal, can specifically include：

In the embodiment of the present invention, the embodiment of the present invention is illustrated so that first updates as an example, for example：It is false If cryptogram is " weare ", the corresponding key press time of the cryptogram is t1, t2, t3, t4, t5； Assuming that the history input number of times of the cryptogram got according to above-mentioned cryptogram " weare " is Count, history average is:Vt6, vt7, vt8, vt9, vt10；History variance is：Var6, var7, Var8, var9, var10；Then above-mentioned cryptogram is that the average after " weare " updates can be：

Vt1=(vt6*count+t1)/(count+1)；

Vt2=(vt7*count+t1)/(count+1)；

Vt3=(vt7*count+t1)/(count+1)；

Vt4=(vt7*count+t1)/(count+1)；

Vt5=(vt7*count+t1)/(count+1)；

Above-mentioned cryptogram is that the variance after " weare " updates can be：

Var1=(var6*count+ (t1-vt1) * (t1-vt1))/(count+1)；

Var2=(var7*count+ (t2-vt2) * (t2-vt2))/(count+1)；

Var3=(var8*count+ (t3-vt3) * (t3-vt3))/(count+1)；

Var4=(var9*count+ (t4-vt4) * (t4-vt4))/(count+1)；

Var5=(var10*count+ (t5-vt5) * (t5-vt5))/(count+1)；

Wherein, above-mentioned vt1 to vt5 represents that above-mentioned cryptogram is the key press time of " weare " correspondence button Average after renewal, above-mentioned Var1 to Var5 represents that above-mentioned cryptogram is " weare " correspondence button Key press time update after variance, because the parameter in gauss hybrid models has updated, thus on State gauss hybrid models and update completion.

It should be noted that the described second process for updating is with reference to above-mentioned first renewal process, this hair Bright embodiment is repeated no more to the second process for updating.

It should be noted that for embodiment of the method, in order to be briefly described, therefore it is all expressed as into one The combination of actions of series, but those skilled in the art should know, and the embodiment of the present invention does not receive institute The limitation of the sequence of movement of description, because according to the embodiment of the present invention, some steps can use other Order is carried out simultaneously.Secondly, those skilled in the art should also know, described in the specification Embodiment belong to preferred embodiment, the involved action not necessarily embodiment of the present invention is musted Must.

Device embodiment

Reference picture 9, shows a kind of structured flowchart of user authentication device embodiment of the invention, specifically may be used With including such as lower module：Acquisition module 901, the first authentication module 902 and the second authentication module 903； Wherein,

Above-mentioned acquisition module 901, can be used for the input behavior of the cryptogram for gathering active user's input Feature；

First authentication module 902, can be used for carrying out first to the cryptogram of active user input Certification；

Second authentication module 903, can be used for after first certification passes through, according to the line of input Be characterized corresponding characteristic model carries out the second certification to the input behavior feature, corresponding to obtain Second authentication result, when second certification passes through, determines active user's certification success, described the Two certifications are obstructed out-of-date, determine active user's authentification failure；Wherein, the characteristic model is according to institute The validated user of cryptogram corresponding account is stated in the input cryptogram or the cryptogram pair The model for answering history input behavior feature during cryptogram fragment to set up.

In a kind of alternative embodiment of the invention, the characteristic model can according to the cryptogram An input behavior feature and the model set up, then second authentication module 903, can specifically wrap Include：

First confidence calculations submodule, can be used for being input into the input behavior feature to corresponding spy Levy in model, corresponding first confidence level is exported by the characteristic model；

First fusion submodule, can be used for according to corresponding first confidence level of all input behavior features, Obtain the confidence level of active user；

First judging submodule, can be used for, according to the confidence level of active user, judging second certification Whether pass through.

In another alternative embodiment of the invention, institute's input model can according to the password Multiple input behavior features of text and the model set up, then second authentication module 903, specifically may be used To include：

Confidence level submodule is calculated, can be used for being input into multiple input behavior features to corresponding character modules In type, the confidence level of active user is exported by the characteristic model；

Second judging submodule, can be used for, according to the confidence level of active user, judging second certification Whether pass through.

In another alternative embodiment of the invention, above-mentioned then described second authentication module 903, specifically Can include：

Cutting submodule, can be used for for the cryptogram being cut into multiple cryptogram fragments；Wherein, The cryptogram fragment has corresponding input behavior characteristic fragment；

Second confidence calculations submodule, can be used for the corresponding input behavior of the cryptogram fragment Characteristic fragment is input into corresponding characteristic model, and corresponding second confidence is exported by the characteristic model Degree；

Second fusion submodule, can be used for according to corresponding second confidence level of all cryptogram fragments, Obtain the confidence level of active user；

3rd judging submodule, can be used for, according to the confidence level of active user, judging second certification Whether pass through.

In a kind of alternative embodiment of the invention, the length of above-mentioned each cryptogram fragment is identical, or Person, the length of each cryptogram fragment is different；Wherein, the length of each cryptogram fragment is not Together, including：The length of the cryptogram fragment according to it is described it is vertical sequentially.

In another alternative embodiment of the invention, the user authentication device can also specifically include：

First searching modul, can be used for searching the index with the cryptogram as major key；Wherein, institute Index is stated with the cryptogram as major key, the storage location with the characteristic model is as storage value；

First acquisition module, can be used for according to the storage location for indexing and obtaining characteristic model, and obtain Take corresponding characteristic model.

Second searching modul, can be used for when the index with the cryptogram as major key is not found, Search the first index with the depression of order text of the cryptogram as major key；

Second acquisition module, can be used for being obtained according to the described first index the storage location of characteristic model, And obtain corresponding characteristic model.

In another alternative embodiment of the invention, the characteristic model is gauss hybrid models, then institute Stating user authentication device specifically can be by the characteristic model as described in lower module foundation：

Set up characteristic model module, can be used for collecting the corresponding history input behavior feature of cryptogram, And/or, the corresponding history input behavior characteristic fragment of cryptogram fragment, calculate history input behavior feature And/or the average and variance of history input behavior characteristic fragment, and set up with the average and variance as parameter Gauss hybrid models.

Update module, can be used for when second certification passes through, according to active user correspondence institute The input behavior feature for stating cryptogram carries out first to the corresponding characteristic model of the input behavior feature Update, or, the input behavior characteristic fragment according to cryptogram fragment is to the input behavior feature piece The corresponding characteristic model of section carries out the second renewal.

In a kind of alternative embodiment of the invention, the characteristic model is gauss hybrid models, then described Update module, can specifically include：

First updates submodule, and history input number of times, the history that can be used for obtaining the cryptogram are defeated Enter first average and first variance of behavioural characteristic, and according to the active user correspondence cryptogram Input behavior feature, first average and first variance are updated；

Second updates submodule, can be used for obtaining the history input number of times of the cryptogram fragment, goes through Second average and second variance of history input behavior characteristic fragment, and according to active user correspondence The input behavior characteristic fragment of cryptogram fragment, is carried out more to second average and second variance Newly.

In another alternative embodiment of the invention, the input behavior feature can specifically include：Press At least one in key time, button coordinate and key press.

For device embodiment, because it is substantially similar to embodiment of the method, so the ratio of description Relatively simple, the relevent part can refer to the partial explaination of embodiments of method.

Each embodiment in this specification is described by the way of progressive, and each embodiment is stressed Be all difference with other embodiment, between each embodiment identical similar part mutually referring to .

On the device in above-described embodiment, wherein modules perform the concrete mode for operating and are having It has been described in detail in the embodiment for closing the method, explanation will be not set forth in detail herein.

Figure 10 is a kind of frame of the device 1000 for user authentication according to an exemplary embodiment Figure.For example, device 1000 can be mobile phone, computer, digital broadcast terminal, information receiving and transmitting Equipment, game console, tablet device, Medical Devices, body-building equipment, personal digital assistant etc..

Reference picture 10, device 1000 can include following one or more assemblies：Processing assembly 1002, Memory 1004, power supply module 1006, multimedia groupware 1008, audio-frequency assembly 1010, input/output (I/O) interface 1012, sensor cluster 1014, and communication component 1016.

The integrated operation of the usual control device 1000 of processing assembly 1002, such as with display, call, Data communication, camera operation and the associated operation of record operation.Treatment element 1002 can include one Individual or multiple processors 1020 carry out execute instruction, to complete all or part of step of above-mentioned method. Additionally, processing assembly 1002 can include one or more modules, be easy to processing assembly 1002 and other Interaction between component.For example, processing component 1002 can include multi-media module, to facilitate many matchmakers Interaction between body component 1008 and processing assembly 1002.

Memory 1004 is configured as storing various types of data supporting the operation in equipment 1000. The example of these data includes the instruction for any application program or method operated on device 1000, Contact data, telephone book data, message, picture, video etc..Memory 1004 can be by any The volatibility or non-volatile memory device or combinations thereof of type realize that such as static random-access is deposited Reservoir (SRAM), Electrically Erasable Read Only Memory (EEPROM), erasable programmable is only Reading memory (EPROM), programmable read only memory (PROM), read-only storage (ROM), Magnetic memory, flash memory, disk or CD.

Power supply module 1006 provides electric power for the various assemblies of device 1000.Power supply module 1006 can be with Including power-supply management system, one or more power supplys, and other generated with device 1000, manage and point With the component that electric power is associated.

Multimedia groupware 1008 is included in one output interface of offer between described device 1000 and user Screen.In certain embodiments, screen can include liquid crystal display (LCD) and touch panel (TP). If screen includes touch panel, screen may be implemented as touch-screen, to receive the input from user Signal.Touch panel includes one or more touch sensors with sensing touch, slip and touch panel Gesture.The touch sensor can not only sensing touch or sliding action border, but also detect The duration related to the touch or slide and pressure.In certain embodiments, multimedia group Part 1008 includes a front camera and/or rear camera.When equipment 1000 is in operator scheme, During such as screening-mode or video mode, front camera and/or rear camera can receive outside many matchmakers Volume data.Each front camera and rear camera can be a fixed optical lens systems or tool There is focusing and optical zoom capabilities.

Audio-frequency assembly 1010 is configured as output and/or input audio signal.For example, audio-frequency assembly 1010 Including a microphone (MIC), when device 1000 is in operator scheme, such as call model, record mould When formula and speech recognition mode, microphone is configured as receiving external audio signal.The audio letter for being received Number can be further stored in memory 1004 or be sent via communication component 1016.In some implementations In example, audio-frequency assembly 1010 also includes a loudspeaker, for exports audio signal.

I/O interfaces 1012 are above-mentioned outer to provide interface between processing assembly 1002 and peripheral interface module It can be keyboard to enclose interface module, click wheel, button etc..These buttons may include but be not limited to：Homepage Button, volume button, start button and locking press button.

Sensor cluster 1014 includes one or more sensors, for providing each side for device 1000 The state estimation in face.For example, sensor cluster 1014 can detect equipment 1000 beat opening/closing shape State, the relative positioning of component, such as described component is the display and keypad of device 1000, sensing Device assembly 1014 can be changed with the position of 1,000 1 components of detection means 1000 or device, user It is presence or absence of with what device 1000 was contacted, the orientation of device 1000 or acceleration/deceleration and device 1000 Temperature change.Sensor cluster 1014 can include proximity transducer, be configured to appoint no The presence of object nearby is detected during what physical contact.Sensor cluster 1014 can also include light sensing Device, such as CMOS or ccd image sensor, for being used in imaging applications.In some embodiments In, the sensor cluster 1014 can also include acceleration transducer, gyro sensor, magnetic sensing Device, pressure sensor or temperature sensor.

Communication component 1016 is configured to facilitate wired or wireless way between device 1000 and other equipment Communication.Device 1000 can access the wireless network based on communication standard, such as WiFi, 2G or 3G, Or combinations thereof.In one exemplary embodiment, communication component 1016 is received via broadcast channel Broadcast singal or broadcast related information from external broadcasting management system.In an exemplary embodiment In, the communication component 1016 also includes near-field communication (NFC) module, to promote junction service. For example, radio frequency identification (RFID) technology, Infrared Data Association (IrDA) can be based in NFC module Technology, ultra wide band (UWB) technology, bluetooth (BT) technology and other technologies are realized.

In the exemplary embodiment, device 1000 can be by one or more application specific integrated circuits (ASIC), digital signal processor (DSP), digital signal processing appts (DSPD), programmable patrol Collect device (PLD), field programmable gate array (FPGA), controller, microcontroller, microprocessor Device or other electronic components are realized, for performing the above method.

In the exemplary embodiment, a kind of non-transitory computer-readable storage including instructing is additionally provided Medium, such as, including the memory 1004 for instructing, above-mentioned instruction can be by the processor 1020 of device 1000 Perform to complete the above method.For example, the non-transitorycomputer readable storage medium can be ROM, random access memory (RAM), CD-ROM, tape, floppy disk and optical data storage devices Deng.

A kind of non-transitorycomputer readable storage medium, when the instruction in the storage medium is by moving During the computing device of terminal so that mobile terminal is able to carry out a kind of user authen method, the side Method includes：The input behavior feature of the cryptogram of collection active user's input；It is defeated to the active user The cryptogram for entering carries out the first certification；It is special according to the input behavior after first certification passes through Levy corresponding characteristic model carries out the second certification to the input behavior feature, to obtain corresponding second Authentication result, when second certification passes through, determines active user's certification success, recognizes described second Demonstrate,prove obstructed out-of-date, determine active user's authentification failure；Wherein, the characteristic model is according to described close The validated user of code text corresponding account is close in the input cryptogram or cryptogram correspondence The model that history input behavior feature during code text fragments is set up.

Figure 11 is the structural representation of server in the embodiment of the present invention.The server 1900 can be because of configuration Or performance is different and the larger difference of producing ratio, can include one or more central processing units (central processing units, CPU) 1922 (for example, one or more processors) and The storage medium of memory 1932, one or more storage application programs 1942 or data 1944 1930 (such as one or more mass memory units).Wherein, memory 1932 and storage medium 1930 can be of short duration storage or persistently storage.The program stored in storage medium 1930 can include one Individual or more than one module (diagram is not marked), each module can include to server in it is a series of Command operating.Further, central processing unit 1922 could be arranged to be communicated with storage medium 1930, Series of instructions operation in performing storage medium 1930 on server 1900.

Server 1900 can also include one or more power supplys 1926, and one or more have Line or radio network interface 1950, one or more input/output interfaces 1958, one or one With upper keyboard 1956, and/or, one or more operating systems 1941, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM etc..

Those skilled in the art will readily occur to this after considering specification and putting into practice invention disclosed herein Other embodiments of invention.It is contemplated that covering any modification of the invention, purposes or adaptability Change, these modifications, purposes or adaptations follow general principle of the invention and including this public affairs Open undocumented common knowledge or conventional techniques in the art.Description and embodiments only by It is considered as exemplary, true scope and spirit of the invention are pointed out by following claim.

It should be appreciated that the invention is not limited in the accurate knot being described above and be shown in the drawings Structure, and can without departing from the scope carry out various modifications and changes.The scope of the present invention is only by appended Claim limit

Presently preferred embodiments of the present invention is the foregoing is only, is not intended to limit the invention, it is all in the present invention Spirit and principle within, any modification, equivalent substitution and improvements made etc. should be included in this hair Within bright protection domain.

Above to a kind of user authen method provided by the present invention, a kind of user authentication device and a kind of use In the device of user authentication, it is described in detail, specific case used herein is to original of the invention Reason and implementation method are set forth, and the explanation of above example is only intended to help and understands side of the invention Method and its core concept；Simultaneously for those of ordinary skill in the art, according to thought of the invention, Will change in specific embodiments and applications, in sum, this specification content is not It is interpreted as limitation of the present invention.

Claims

1. a kind of user authen method, it is characterised in that including：

Cryptogram to active user input carries out the first certification；

2. authentication method according to claim 1, it is characterised in that the characteristic model is root The model set up according to an input behavior feature of the cryptogram, then it is described according to the line of input The step of being characterized corresponding characteristic model and carry out the second certification to the input behavior feature, including：

3. authentication method according to claim 1, it is characterised in that the input model is root The model set up according to multiple input behavior features of the cryptogram, then it is described according to the line of input The step of being characterized corresponding characteristic model and carry out the second certification to the input behavior feature, including：

4. authentication method according to claim 1, it is characterised in that described according to the input The step of corresponding characteristic model of behavioural characteristic carries out the second certification to the input behavior feature, bag Include：

5. method according to claim 4, it is characterised in that the length of each cryptogram fragment Degree is identical, or, the length of each cryptogram fragment is different；

6. method according to claim 1, it is characterised in that described according to the line of input Before corresponding characteristic model is characterized the step of carrying out the second certification to the input behavior feature, institute Stating method also includes：

7. method according to claim 6, it is characterised in that methods described also includes：

8. according to any described method in claim 1 to 7, it is characterised in that the character modules Type is gauss hybrid models, then set up the characteristic model as follows：

9. according to any described method in claim 1 to 7, it is characterised in that methods described is also Including：

10. method according to claim 9, it is characterised in that the characteristic model is mixed for Gauss Matched moulds type, then the step of first renewal, including：

The step of second renewal, including：

11. according to any described method in claim 1 to 7, it is characterised in that the line of input Be characterized including：At least one in key press time, button coordinate and key press.

A kind of 12. user authentication devices, it is characterised in that including：

13. a kind of devices for user authentication, it is characterised in that include memory, Yi Jiyi Individual or more than one program, one of them or more than one program storage in memory, and It is configured to that one by one or more than one computing device or more than one program bag contains Instruction for carrying out following operation：

Cryptogram to active user input carries out the first certification；