CN106897412A - A kind of method and apparatus for recommending associated application based on intended application - Google Patents

Abstract

The invention provides a kind of method and apparatus for recommending associated application based on intended application.Methods described includes：The application being associated with intended application is searched for according to label；One or more appropriate parameters are selected to determine matching degree of the searched for associated application for the intended application；The associated application descending searched for is arranged based on the matching degree size and order is recommended to user.

Description

A kind of method and apparatus for recommending associated application based on intended application

Technical field

The present invention relates to technical field of information processing, associated application is recommended based on intended application in particular to one kind Method and apparatus.

Background technology

With the fast development of Internet technology and intelligent mobile terminal technology, the work(much realized on computer terminals (can for example do shopping, read) can also realize on intelligent mobile terminal, for example, use smart mobile phone or panel computer etc..Separately Outward, the realization of these functions needs to install corresponding application program on intelligent mobile terminal.For example, shopping online is, it is necessary to pacify Dress such as Taobao's client, listens music to need to install music player client etc..Thus, many software companys provide application Shop or application market, such as pea pods or PP assistant etc..User can be opened using shop or application market, so as to Various application programs required for enough fast searchs and download, including audio-visual broadcast message class, system tool class, the social class of communication, net Upper shopping class, reading class etc., certainly can be with amusement and recreation class application program such as download games (APP).

In order to constantly lift good experience sense of the user using application shop or application market, current developer develops The function that many convenient users are used, one of them is recommendation function, i.e., recommend some to apply to user, to help user to find More applications interested.A kind of common recommendation exhibition method is that associated answering is recommended in the application currently clicked on according to user With " people for having downloaded * * * can also download " shown in " everybody also downloads ", Figure 1B for example shown in Figure 1A.This kind of scene is pushed away It is to recommend a collection of associated application to user according to current application to recommend logic.Traditional recommendation method was cooperateed with using label Filtering method, i.e.,：Limit applying having same label with intended application of recommending first, then by the download of user, browse, Behavioral data is installed etc. and sets up the user behavior space vector of each application, finally according to cosine coefficient (or the outstanding German number of card, Pierre Gloomy coefficient etc.) Similarity value for recommending application and intended application is calculated, a collection of application of similarity ranking foremost is taken as pushing away Recommend candidate's application.

But above-mentioned recommendation method Shortcomings in scene is recommended in association application are：Its main thought is to download Which application the people of intended application can also download, it is considered to it is important that leaving for finding which application is related from the behavior of user Property it is higher, but lack consideration to application quality in itself, the application of so recommendation is probably " fair without, foul within ", its Reason is：Applying for some poor qualities is made very well in packaging, and therefore many users are attracted by it and produce click behavior, This can cause existing recommendation method to think that this recommends out using very popular with users it, and substantial these applications Experience sense is very poor.In brief, the shortcoming of above-mentioned existing recommendation method is that easily the very poor application of experience sense is recommended User, causes the experience sense of application shop that user installs to it or application market to be deteriorated.

The content of the invention

It is above-mentioned to improve it is an object of the invention to a kind of method and apparatus for recommending associated application based on intended application Problem.

A kind of method for recommending associated application based on intended application is the embodiment of the invention provides, it includes：

The application being associated with intended application is searched for according to label；

One or more appropriate parameters are selected to determine matching of the searched for associated application for the intended application Degree；

The associated application descending searched for is arranged based on the matching degree size and order is recommended to user.

Preferably, in the step of according to label to search for the application being associated with intended application, searching out has and mesh Mark the application of the label same label of application.

The embodiment of the present invention additionally provides a kind of device based on intended application recommendation associated application, and it includes：

Search unit, for searching for the application being associated with intended application according to label；

Matching degree determining unit, searched for associated application pair is determined for selecting one or more appropriate parameters In the matching degree of the intended application；

Recommendation unit, for based on the matching degree size to searched for associated application descending arrangement and order to User recommends.

Preferably, the search unit is used to search out the application with the label same label of intended application.

Wherein, described one or more parameters include：Temperature of the associated application relative to intended application, scoring matter One of amount, clicking rate and conversion ratio or any 2 parameter combinations or any 3 parameter combinations or 4 parameter combinations or Person's other specification.

Wherein, 1) associated application is calculated as follows relative to the temperature of intended application：

Wherein hot (j) represents temperatures of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Install (i, j) represents whether the user i of installation targets application is mounted with associated application j, value 0 or 1,0 table Show and associated application j is not installed, 1 expression is mounted with associated application j；

2) associated application is as follows relative to the scoring Mass Calculation of intended application：

Wherein evl (j) represents scoring quality of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Evaluate (i, j) represents the user i of installation targets application to the evaluation score of associated application j, its value for- Integer between 1 or [0,5], represents that user does not provide evaluation score when value is -1.

Function if () is 0 in variable evaluate (i, j)=- 1, and otherwise if () is 1；

Function if ' () is 0 in variable evaluate (i, j)=- 1, and otherwise if ' () is evaluate (i, j)；

3) associated application is calculated as follows relative to the clicking rate of intended application：

Wherein ctr (j) represents clicking rates of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Click (i, j) represents whether the associated application j to showing to it produces click for the user i of installation targets application Behavior, click (i, j) value is -1,0,1, wherein -1 expression did not show associated application j to user i, 0 represents to user i Illustrate associated application j but user i does not click on associated application j, 1 represents to user i displaying associated application j and user i points Associated application j is hit；

Function if () is 0 in variable click (i, j)=- 1, and otherwise if () is 1；

Function if ' () is 0 in variable click (i, j)=- 1, and otherwise if ' () is click (i, j)；

4) associated application is calculated as follows relative to the conversion ratio of intended application：

Wherein dtr (j) represents conversion ratios of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Down (i, j) represents whether the associated application j to showing to it is produced downloads row for the user i of installation targets application For down (i, j) value is -1,0,1, wherein -1 expression did not show associated application j to user i, 0 represents to user's i exhibitions Associated application j is shown but user i has not downloaded associated application j, 1 has represented to user i displaying associated application j and user i downloads Associated application j；

Function if () is 0 in variable down (i, j)=- 1, and otherwise if () is 1；

Function if ' () is 0 in variable down (i, j)=- 1, and otherwise if ' () is down (i, j).

Wherein it is determined that the associated application searched for is as follows for the method for the matching degree of the intended application：

Wherein：Fit (j) represents matching degrees of the associated application j relative to intended application；

Avg (hot (j)) represents the average value of the temperature of all associated applications being retrieved；

Avg (evl (j)) represents the average value of the scoring quality of all associated applications being retrieved；

Avg (ctr (j)) represents the average value of the clicking rate of all associated applications being retrieved；

Avg (dtr (j)) represents the average value of the conversion ratio of all associated applications being retrieved；

α, β, γ and θ are the weights for adjusting each factor, wherein alpha+beta+γ+θ=1, and α, β, γ and θ ∈ [0,1], Determine to be calculated by those parameter indexs searched for associated application for the mesh by taking α, β, γ and θ different value Mark the matching degree of application.

The method and apparatus for being recommended associated application based on intended application of the invention, are gone out by label search first The application being associated with intended application, its interrelational form is that the application being retrieved has identical label with intended application, so Consider afterwards the associated application that is retrieved relative to the temperature of intended application, clicking rate, conversion ratio, scoring quality or other The factors such as parameter, are used to weigh the matching degree of associated application and intended application, finally according to the size preferential recommendation of matching degree With the big associated application of degree.With the application that this scheme recommends out, similarity factor is both considered, it is contemplated that recommending application Quality, improve the deficiency that prior art easily recommends experience sense very poor application user, improve Consumer's Experience.

Brief description of the drawings

Figure 1A and 1B are the schematic sectional drawings for recommending associated application according to the displaying of the recommendation method of prior art；

Fig. 2 is the indicative flowchart of the method for recommending associated application based on intended application of the invention；

Fig. 3 be it is exemplary show to be shown in application market with 2 sectional drawings of the application of label.

Fig. 4 is the schematic block diagram of the device for recommending associated application based on intended application of the invention.

Specific embodiment

Below in conjunction with the embodiment of the present invention and accompanying drawing, the technical scheme in the embodiment of the present invention is carried out clear, complete Ground description, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.Generally herein The component of the embodiment of the present invention described and illustrated in place's accompanying drawing can be arranged and designed with a variety of configurations.Therefore, The detailed description of the embodiments of the invention to providing in the accompanying drawings is not intended to limit the model of claimed invention below Enclose, but be merely representative of selected embodiment of the invention.Based on embodiments of the invention, those skilled in the art are not making The every other embodiment obtained on the premise of creative work, belongs to the scope of protection of the invention.

The method for recommending associated application based on intended application provided by the present invention is described in detail below by embodiment And device.

Fig. 2 is a kind of indicative flowchart of the method for recommending associated application based on intended application.As shown in Fig. 2 this The method based on intended application recommendation associated application of invention is comprised the following steps：

S1：The application being associated with intended application is searched for according to label.

Generally, all there is label, mark using the various application programs (referred to as using) provided in shop or application market The effect of label is classification or the content for identifying various application programs, is easy to user to search.At present, in application market or shop is applied In each application can include at least one application label, as shown in figure 3, Fig. 3 shows entertainment applications " happy bucket ground everyday It is main " 2 labels are included, 1 label shows that its content for identifying the application is " fighting landlord ", and another 1 label shows that its mark should The classification of application is " playing cards ".

Therefore in this step, can be with according to label come the implementation method for searching for the application being associated with intended application Search out with the application with the label same label of intended application as associated application, so-called intended application refers to just user The current application clicked on.As a example by shown in Fig. 3, all applications that outgoing label content is " fighting landlord " had both been may search for, it is also possible to It is to seek out all applications that label substance is " playing cards ", and includes comprising appointing according to the associated application that is retrieved of the present invention The application of meaning same label, that is, the associated application being retrieved includes all applications for " fighting landlord " with label substance, It is this two parts of all applications of " playing cards " with label substance.

The application for having same label with intended application is retrieved first, so as to will first recommend application to be defined to generic phase Associated application, can apply recommendation has certain similarity with intended application, can also reduce operand below.

S2：One or more appropriate parameters are selected to determine searched for associated application for the intended application Matching degree.

As being described above, application is only recommended in the way of similarity, easily the very poor application of experience sense is pushed away Recommend out.This is accomplished by considering the quality of the associated application that previous step is retrieved.Can be counted by some parameter indexs Searched for matching degree of the associated application for the intended application is calculated, value according to matching degree weighs associated application Quality.

One or more appropriate parameters described here include：The associated application relative to intended application temperature, comment One of sub-prime amount, clicking rate and conversion ratio or any 2 parameter combinations therein or any 3 parameter combinations or this 4 ginsengs Array is closed or other specification.

Specifically, when 1 parameter is selected, the parameter can be heat of the associated application relative to intended application Degree, scoring quality, clicking rate or conversion ratio, or other specification.

Wherein, the temperature is exactly the probability for being mounted with some associated application of the user installation of intended application；

The scoring quality is exactly mean opinion score of the user to some associated application for being mounted with intended application；

The clicking rate is exactly to be mounted with that the user of intended application clicks on the number of times of some associated application and shown with it Show the ratio between number of times；

The conversion ratio is exactly to be mounted with that the user of intended application downloads the number of times of some related application and is shown with it The ratio between number of times.

1) associated application is calculated as follows relative to the temperature of intended application：

Wherein hot (j) represents temperatures of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Install (i, j) represents whether the user i of installation targets application is mounted with associated application j, value 0 or 1,0 table Show and associated application j is not installed, 1 expression is mounted with associated application j.

The temperature that can be drawn by the formula is mounted with some associated application of the user installation of intended application j Probability.

Function if () and if ' () can be used when calculating scoring quality, clicking rate and conversion ratio below, wherein

The computing of function if () is as follows：

The computing of function if ' () is as follows：

2) associated application is as follows relative to the scoring Mass Calculation of intended application：

Wherein evl (j) represents scoring quality of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Evaluate (i, j) represents the user i of installation targets application to the evaluation score of associated application j, its value for- Integer between 1 or [0,5], represents that user does not provide evaluation score when value is -1.

Function if (evaluate (i, j)) is 0 at evaluate (i, j)=- 1, and otherwise if (evaluate (i, j)) is 1；

Function if ' (evaluate (i, j)) is 0, otherwise if ' (evaluate (i, j)) at evaluate (i, j)=- 1 It is evaluate (i, j).

The scoring quality that can be drawn by the formula is exactly the user for being mounted with intended application to some associated application j Mean opinion score, i.e., in all users for being mounted with intended application, do not consider not carry out a certain associated application j The user data of evaluation score, mean opinion score of remaining user for being drawn to some associated application j.

3) associated application is calculated as follows relative to the clicking rate of intended application：

Wherein ctr (j) represents clicking rates of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Click (i, j) represents whether the associated application j to showing to it produces click for the user i of installation targets application Behavior, click (i, j) value is -1,0,1, wherein -1 expression did not show associated application j to user i, 0 represents to user i Illustrate associated application j but user i does not click on associated application j, 1 represents to user i displaying associated application j and user i points Associated application j is hit；

Function if (click (i, j)) is 0 at click (i, j)=- 1, and otherwise if (click (i, j)) is 1；

Function of ' (click (i, j)) is 0 at click (i, j)=- 1, and otherwise if ' (click (i, j)) is click (i,j)。

The clicking rate that can be drawn by the formula is exactly that the user for being mounted with intended application clicks on some associated application The ratio between number of times and its shown number of times, i.e., in all users for being mounted with intended application, do not consider not show to user i The user data of associated application j is crossed, the clicking rate to some associated application j for being drawn.

4) associated application is calculated as follows relative to the conversion ratio of intended application：

Wherein dtr (j) represents conversion ratios of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Down (i, j) represents whether the associated application j to showing to it is produced downloads row for the user i of installation targets application For down (i, j) value is -1,0,1, wherein -1 expression did not show associated application j to user i, 0 represents to user's i exhibitions Associated application j is shown but user i has not downloaded associated application j, 1 has represented to user i displaying associated application j and user i downloads Associated application j；

Function if (down (i, j)) is 0 at down (i, j)=- 1, and otherwise if (down (i, j)) is 1；

Function if ' (down (i, j)) is 0 at down (i, j)=- 1, and otherwise if ' (down (i, j)) is down (i, j).

The conversion ratio that can be drawn by the formula is mounted with that the user of intended application downloads some related application The ratio between number of times and its shown number of times, i.e., in all users for being mounted with intended application, do not consider not show to user i The user data of associated application j, the conversion ratio to some associated application j for being drawn.

So, one or more appropriate parameters are selected to determine searched for associated application for the intended application Matching degree method it is as follows：

Wherein：Fit (j) represents matching degrees of the associated application j relative to intended application；

Avg (hot (j)) represents the average value of the temperature of all associated applications being retrieved；

Avg (evl (j)) represents the average value of the scoring quality of all associated applications being retrieved；

Avg (ctr (j)) represents the average value of the clicking rate of all associated applications being retrieved；

Avg (dtr (j)) represents the average value of the conversion ratio of all associated applications being retrieved；

By the value of each factor (temperature, scoring quality, clicking rate and conversion ratio) divided by the average value of the factor purpose It is that standardization is done to the factor value.

α, β, γ and θ are the weights for adjusting each factor, wherein alpha+beta+γ+θ=1, and α, β, γ and θ ∈ [0,1], Can determine to be calculated by those parameter indexs searched for associated application for institute by taking α, β, γ and θ different value State the matching degree of intended application.For example, when α=1, β=0, γ=0 and θ=0, then matching degree is calculated by temperature index, When α=0, β=1, γ=0 and θ=0, then matching degree is calculated by scoring quality index, by that analogy, when α=0, β= 0th, γ=0.5 and θ=0.5 (γ and θ can also negated 0 other values, and when meeting γ+θ=1), then by clicking rate Matching degree is calculated with conversion ratio index, it is also possible to which (α and β can also negated 0 to allow α=0.5, β=0.5, γ=0 and θ=0 Other values, and meet alpha+beta=1) or α=0, β=0.5, γ=0.5 and θ=0 (β and γ can also negated 0 other Value, and meet β+γ=1), other can also be chosen and combined；For another example, α=0.5, β=0.3, γ=0.2 and θ can be allowed =0 (α, β and γ can also negated 0 other values, and alpha+beta+γ=1) is met, now by temperature, scoring quality and point Hit rate index to calculate matching degree, it is also possible to other combinations of Xuanzhou District；If this 4 kinds of factors considered, alpha+beta+γ+θ simultaneously =1, α, β, γ and θ ∈ (0,1], i.e., α, β, γ and θ are not 0；If this 4 kinds of factors are of equal importance, can take α=β= γ=θ=0.25.Can also be drawn by these examples, can select the associated application relative to intended application temperature, One of scoring quality, clicking rate and conversion ratio or any 2 parameter combinations therein or any 3 parameter combinations or this 4 Parameter combination determines matching degree of the searched for associated application for the intended application.

S3：The associated application descending searched for is arranged based on the matching degree size and order is recommended to user.

Descending arrangement is done from big to small according to matching angle value to the application being associated with intended application searched for, then Result is shown successively in the page as the correlation recommendation list of application of intended application by tab sequential.

The method for being recommended associated application based on intended application of the invention, is gone out and target by label search first Using associated application, its interrelational form is that the application being retrieved has identical label with intended application, then comprehensive Temperature, clicking rate, conversion ratio, scoring quality or other specification etc. of the associated application that is retrieved of consideration relative to intended application Factor, is to consider application to user using clicking rate factor wherein the use of temperature factor being the similitude from from the point of view of user Attraction, be the quality for considering application using conversion ratio and scoring quality, thus weigh associated application and intended application With degree, finally according to the big associated application of the size preferential recommendation matching degree of matching degree.With the application that this scheme recommends out, Both considered similarity factor, it is contemplated that recommend application quality, improve prior art easily experience sense it is very poor should With the deficiency for recommending user, Consumer's Experience is improve.

Fig. 4 is the schematic block diagram of the device for recommending associated application based on intended application of the invention.As shown in figure 4, Device based on intended application recommendation associated application of the invention includes：

Search unit, for searching for the application being associated with intended application according to label；

Matching degree determining unit, searched for associated application pair is determined for selecting one or more appropriate parameters In the matching degree of the intended application；

Recommendation unit, for based on the matching degree size to searched for associated application descending arrangement and order to User recommends.

Wherein, search unit searches for implementing process and can joining for the application that is associated with intended application according to label See above-mentioned corresponding method and step, i.e., described search unit preferably search out with the label same label of intended application should With.

Matching degree determining unit selects one or more appropriate parameters to determine searched for associated application for described The process that implements of the matching degree of intended application can also be referring to above-mentioned corresponding method and step.Here sketch again.

One or more appropriate parameters described here include：The associated application relative to intended application temperature, comment One of sub-prime amount, clicking rate and conversion ratio or any 2 parameter combinations therein or any 3 parameter combinations or this 4 ginsengs Array is closed or other specification.

Wherein, the temperature is exactly the probability for being mounted with some associated application of the user installation of intended application；

The scoring quality is exactly mean opinion score of the user to some associated application for being mounted with intended application；

The clicking rate is exactly to be mounted with that the user of intended application clicks on the number of times of some associated application and shown with it Show the ratio between number of times；

The conversion ratio is exactly to be mounted with that the user of intended application downloads the number of times of some related application and is shown with it The ratio between number of times.

Described temperature, the computational methods of score quality, clicking rate and conversion ratio are referring to detailed process presented hereinbefore.This In repeat no more.

The computational methods of matching degree are also identical with presented hereinbefore, i.e.,

Wherein：Fit (j) represents matching degrees of the associated application j relative to intended application；

Avg (hot (j)) represents the average value of the temperature of all associated applications being retrieved；

Avg (evl (j)) represents the average value of the scoring quality of all associated applications being retrieved；

Avg (ctr (j)) represents the average value of the clicking rate of all associated applications being retrieved；

Avg (dtr (j)) represents the average value of the conversion ratio of all associated applications being retrieved；

α, β, γ and θ are the weights for adjusting each factor, wherein alpha+beta+γ+θ=1, and α, β, γ and θ ∈ [0,1], Can determine to be calculated by those parameter indexs searched for associated application for institute by taking α, β, γ and θ different value State the matching degree of intended application.

By the value of each factor (temperature, scoring quality, clicking rate and conversion ratio) divided by the average value of the factor purpose It is that standardization is done to the factor value.

Recommendation unit is based on associated application descending of the matching degree size to being searched for and arranges and sequentially to user Recommend.

It is apparent to those skilled in the art that, for convenience and simplicity of description, the device of foregoing description Specific work process, may be referred to the corresponding process in preceding method embodiment, be not repeated herein description.

Device based on intended application recommendation associated application of the invention, is gone out and target by label search first Using associated application, its interrelational form is that the application being retrieved has identical label with intended application, then comprehensive Temperature, clicking rate, conversion ratio, scoring quality or other specification etc. of the associated application that is retrieved of consideration relative to intended application Factor, is to consider application to user using clicking rate factor wherein the use of temperature factor being the similitude from from the point of view of user Attraction, be the quality for considering application using conversion ratio and scoring quality, thus weigh associated application and intended application With degree, finally according to the big associated application of the size preferential recommendation matching degree of matching degree.With the application that this scheme recommends out, Both considered similarity factor, it is contemplated that recommend application quality, improve prior art easily experience sense it is very poor should With the deficiency for recommending user, Consumer's Experience is improve.

The computer program product of the method for recommending associated application based on intended application that the embodiment of the present invention is provided, Computer-readable recording medium including storing program code, the instruction that described program code includes can be used for side before execution Method described in method embodiment, implements and can be found in embodiment of the method, will not be repeated here.

If the function is to realize in the form of SFU software functional unit and as independent production marketing or when using, can be with Storage is in a computer read/write memory medium.Based on such understanding, technical scheme is substantially in other words The part contributed to prior art or the part of the technical scheme can be embodied in the form of software product, the meter Calculation machine software product is stored in a storage medium, including some instructions are used to so that a computer equipment (can be individual People's computer, panel computer, smart mobile phone, server, or network equipment etc.) perform each embodiment methods described of the invention All or part of step.And foregoing storage medium includes：USB flash disk, mobile hard disk, read-only storage (ROM), arbitrary access are deposited Reservoir (RAM), magnetic disc or CD etc. are various can be with the medium of store program codes.

The above, specific embodiment only of the invention, but protection scope of the present invention is not limited thereto, and it is any Those familiar with the art the invention discloses technical scope in, change or replacement can be readily occurred in, should all contain Cover within protection scope of the present invention.Therefore, protection scope of the present invention should be based on the protection scope of the described claims.

Claims (10)

1. it is a kind of based on intended application recommend associated application method, it includes：

The application being associated with intended application is searched for according to label；

One or more appropriate parameters are selected to determine matching degree of the searched for associated application for the intended application；

The associated application descending searched for is arranged based on the matching degree size and order is recommended to user.

2. method according to claim 1, it is characterised in that according to label come search for be associated with intended application should With the step of in, search out the application with the label same label of intended application.

3. method according to claim 1, it is characterised in that described one or more parameters include：The associated application One of temperature, scoring quality, clicking rate and conversion ratio relative to intended application or any 2 parameter combinations or any 3 Parameter combination or 4 parameter combinations or other specification.

4. method according to claim 3, it is characterised in that 1) temperature of the associated application relative to intended application It is calculated as follows：

$hot\left(j\right)=\frac{{\mathrm{\Σ}}_{i=1}^{n}install(i,j)}{n},j=1,2,...,m$

Wherein hot (j) represents temperatures of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Install (i, j) represents whether the user i of installation targets application is mounted with associated application j, value 0 or 1, and 0 represents not Associated application j is installed, 1 expression is mounted with associated application j；

2) associated application is as follows relative to the scoring Mass Calculation of intended application：

$evl\left(j\right)=\frac{{\mathrm{\Σ}}_{i=1}^n{\mathrm{if}}^{\′}\left(evaluate\right(i,j\left)\right)}{{\mathrm{\Σ}}_{i=1}^{n}if\left(evaluate\right(i,j\left)\right)},j=1,2,...,m$

Wherein evl (j) represents scoring quality of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Evaluate (i, j) represents the user i of installation targets application to the evaluation score of associated application j, its value be -1 or Integer between [0,5], represents that user does not provide evaluation score when value is -1.

Function if () is 0 in variable evaluate (i, j)=- 1, and otherwise if () is 1；

Function if ' () is 0 in variable evaluate (i, j)=- 1, and otherwise if ' () is evaluate (i, j)；

3) associated application is calculated as follows relative to the clicking rate of intended application：

$ctr\left(j\right)=\frac{{\Σ}_{i=1}^n{\mathrm{if}}^{\′}\left(click\right(i,j\left)\right)}{{\Σ}_{i=1}^{n}if\left(click\right(i,j\left)\right)},j=1,2,\mathrm{...},m$

Wherein ctr (j) represents clicking rates of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Click (i, j) represents whether the associated application j to showing to it produces click behavior for the user i of installation targets application, Click (i, j) value is -1,0,1, wherein -1 expression did not show associated application j to user i, 0 represents to user i displayings Associated application j but user i do not click on associated application j, and 1 to represent and show associated application j to user i and user i is clicked Associated application j；

Function if () is 0 in variable click (i, j)=- 1, and otherwise if () is 1；

Function if ' () is 0 in variable click (i, j)=- 1, and otherwise if ' () is click (i, j)；

4) associated application is calculated as follows relative to the conversion ratio of intended application：

$dtr\left(j\right)=\frac{{\Σ}_{i=1}^n{\mathrm{if}}^{\′}\left(down\right(i,j\left)\right)}{{\Σ}_{i=1}^{n}if\left(down\right(i,j\left)\right)},j=1,2,\mathrm{...},m$

Wherein dtr (j) represents conversion ratios of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Down (i, j) represents whether the associated application j to showing to it produces download behavior for the user i of installation targets application, Down (i, j) value is -1,0,1, represents and is illustrated to user i wherein -1 represents and do not showed associated application j to user i, 0 Associated application j but user i do not download associated application j, and 1 to represent and show associated application j to user i and user i has downloaded this Associated application j；

Function if () is 0 in variable down (i, j)=- 1, and otherwise if () is 1；

Function if ' () is 0 in variable down (i, j)=- 1, and otherwise if ' () is down (i, j).

5. method according to claim 4, it is characterised in that it is determined that the associated application searched for should for the target The method of matching degree is as follows：

$\begin{array}{c}fit\left(j\right)=\mathrm{\α}*\frac{hot}{avg\left(hot\right(j\left)\right)}+\mathrm{\β}*\frac{evl\left(j\right)}{avg\left(evl\right(j\left)\right)}+\mathrm{\γ}*\frac{ctr\left(j\right)}{avg\left(ctr\right(j\left)\right)}+\mathrm{\θ}*\frac{dtr\left(j\right)}{avg\left(dtr\right(j\left)\right)}\\ j=1,2,\mathrm{...},m\end{array}$

Wherein：Fit (j) represents matching degrees of the associated application j relative to intended application；

Avg (hot (j)) represents the average value of the temperature of all associated applications being retrieved；

Avg (evl (j)) represents the average value of the scoring quality of all associated applications being retrieved；

Avg (ctr (j)) represents the average value of the clicking rate of all associated applications being retrieved；

Avg (dtr (j)) represents the average value of the conversion ratio of all associated applications being retrieved；

α, β, γ and θ are the weights for adjusting each factor, wherein alpha+beta+γ+θ=1, and α, β, γ and θ ∈ [0,1], are passed through α, β, γ and θ different value is taken to determine that calculated searched for associated application by those parameter indexs is answered for the target Matching degree.

6. it is a kind of based on intended application recommend associated application device, it includes：

Search unit, for searching for the application being associated with intended application according to label；

Matching degree determining unit, searched for associated application is determined for institute for selecting one or more appropriate parameters State the matching degree of intended application；

Recommendation unit, for being arranged the associated application descending searched for and sequentially to user based on the matching degree size Recommend.

7. device according to claim 6, it is characterised in that the search unit be used to searching out with and intended application Label same label application.

8. device according to claim 6, it is characterised in that described one or more parameters include：The associated application One of temperature relative to intended application, scoring quality, clicking rate and conversion ratio or any 2 parameter combinations or any 3 parameter combinations or 4 parameter combinations or other specification.

9. device according to claim 8, it is characterised in that 1) temperature of the associated application relative to intended application It is calculated as follows：

$hot\left(j\right)=\frac{{\mathrm{\Σ}}_{i=1}^{n}install(i,j)}{n},j=1,2,...,m$

Wherein hot (j) represents temperatures of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Install (i, j) represents whether the user i of installation targets application is mounted with associated application j, value 0 or 1, and 0 represents not Associated application j is installed, 1 expression is mounted with associated application j；

2) associated application is as follows relative to the scoring Mass Calculation of intended application：

$evl\left(j\right)=\frac{{\mathrm{\Σ}}_{i=1}^n{\mathrm{if}}^{\′}\left(evaluate\right(i,j\left)\right)}{{\mathrm{\Σ}}_{i=1}^{n}if\left(evaluate\right(i,j\left)\right)},j=1,2,...,m$

Wherein evl (j) represents scoring quality of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Evaluate (i, j) represents the user i of installation targets application to the evaluation score of associated application j, its value be -1 or Integer between [0,5], represents that user does not provide evaluation score when value is -1.

Function if () is 0 in variable evaluate (i, j)=- 1, and otherwise if () is 1；

Function if ' () is 0 in variable evaluate (i, j)=- 1, and otherwise if ' () is evaluate (i, j)；

3) associated application is calculated as follows relative to the clicking rate of intended application：

$ctr\left(j\right)=\frac{{\Σ}_{i=1}^n{\mathrm{if}}^{\′}\left(click\right(i,j\left)\right)}{{\Σ}_{i=1}^{n}if\left(click\right(i,j\left)\right)},j=1,2,\mathrm{...},m$

Wherein ctr (j) represents clicking rates of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Click (i, j) represents whether the associated application j to showing to it produces click behavior for the user i of installation targets application, Click (i, j) value is -1,0,1, wherein -1 expression did not show associated application j to user i, 0 represents to user i displayings Associated application j but user i do not click on associated application j, and 1 to represent and show associated application j to user i and user i is clicked Associated application j；

Function if () is 0 in variable click (i, j)=- 1, and otherwise if () is 1；

Function if ' () is 0 in variable click (i, j)=- 1, and otherwise if ' () is click (i, j)；

4) associated application is calculated as follows relative to the conversion ratio of intended application：

$dtr\left(j\right)=\frac{{\Σ}_{i=1}^n{\mathrm{if}}^{\′}\left(down\right(i,j\left)\right)}{{\Σ}_{i=1}^{n}if\left(down\right(i,j\left)\right)},j=1,2,\mathrm{...},m$

Wherein dtr (j) represents conversion ratios of the associated application j relative to intended application；

M represents the quantity of the associated application being retrieved；

N represents the quantity of the user of installation targets application；

Down (i, j) represents whether the associated application j to showing to it produces download behavior for the user i of installation targets application, Down (i, j) value is -1,0,1, represents and is illustrated to user i wherein -1 represents and do not showed associated application j to user i, 0 Associated application j but user i do not download associated application j, and 1 to represent and show associated application j to user i and user i has downloaded this Associated application j；

Function if () is 0 in variable down (i, j)=- 1, and otherwise if () is 1；

Function if ' () is 0 in variable down (i, j)=- 1, and otherwise if ' () is down (i, j).

10. device according to claim 9, it is characterised in that it is determined that the associated application searched for should for the target The method of matching degree is as follows：

$\begin{array}{c}fit\left(j\right)=\mathrm{\α}*\frac{hot}{avg\left(hot\right(j\left)\right)}+\mathrm{\β}*\frac{evl\left(j\right)}{avg\left(evl\right(j\left)\right)}+\mathrm{\γ}*\frac{ctr\left(j\right)}{avg\left(ctr\right(j\left)\right)}+\mathrm{\θ}*\frac{dtr\left(j\right)}{avg\left(dtr\right(j\left)\right)}\\ j=1,2,\mathrm{...},m\end{array}$

Wherein：Fit (j) represents matching degrees of the associated application j relative to intended application；

Avg (hot (j)) represents the average value of the temperature of all associated applications being retrieved；

Avg (evl (j)) represents the average value of the scoring quality of all associated applications being retrieved；

Avg (ctr (j)) represents the average value of the clicking rate of all associated applications being retrieved；

Avg (dtr (j)) represents the average value of the conversion ratio of all associated applications being retrieved；

α, β, γ and θ are the weights for adjusting each factor, wherein alpha+beta+γ+θ=1, and α, β, γ and θ ∈ [0,1], are passed through α, β, γ and θ different value is taken to determine that calculated searched for associated application by those parameter indexs is answered for the target Matching degree.