CN103019550A - Real-time display method and system for associated content - Google Patents

Abstract

The invention provides a real-time display method and a real-time display system for associated content. The method comprises the following steps of: acquiring content of a current screen display page of a user terminal in real time according to screen changing action of the user terminal; determining focal content of the screen display page; further judging whether cache information associated with the focal content of the screen display page exists or not; if the cache information exists, acquiring the cache information and displaying the cache information to the user terminal as associated content recommended to the user terminal; otherwise, acquiring other user groups having paid attention to the focal content of the screen display page through real-time computing and other user groups with behaviors similar to the behavior of the user terminal, determining the common focus of attention of the other user groups and displaying the common focus of attention to the user terminal as the associated content recommended to the user terminal. According to the method and the system, contents which are associated with the focus and are probably favored by users are displayed in a changing manner in real time according to different focuses of attention where the users stop at present and to which attention is paid, so that the user experience and the conversion rate of a website are improved.

Description

The real-time exhibiting method of content association and system

Technical field

The present invention relates to the web browsing technology field, more specifically, relate to a kind of real-time exhibiting method of content association and system of recommendation network information.

Background technology

Along with the progress of society and the development of technology, people utilize network to come obtaining information more and more.In this case, by the browsing, search for, click and the behavior such as comment of user, for the user recommends just to have become one of focus content that the company such as portal website, online shopping mall or enterprise pay close attention to the most with their be associated the most content that maybe may like.

Commending system is used on the website of Amazon, according in the past user's buying behavior, and other products of recommending out certain product of purchase may buy simultaneously.Improve the conversion ratio (visit capacity of the action of website expectation/total visit capacity of its electronic bookstore from the commending system that utilizes of Amazon success, this index is used for weighing web site contents to visitor's attraction degree and the effect of publicity of website) afterwards, personalized recommendation system all is widely used in internet product, ecommerce and social network sites at present.In general, proposed algorithm need to be according to a large number of users in the website to the behavior in certain product use procedure (browse, click, comment etc.), in conjunction with relevant proposed algorithm (for example collaborative filtering), find and represent with it to the user content of be associated most (may like).

In recent years, proposed algorithm has obtained continuous retrofit, collaborative filtering for example, this algorithm finds jointly to browse certain page or the user group similar to this user behavior, and from similar users colony, find its joint act and recommend this user, thereby can be more effectively to the user recommend may be interested in good friend perhaps.The crawl of legacy user's behavioral data is divided into recessive data and dominant data, and dominant data owner will comprise evaluation, buy the operation of so direct performance user preferences; Recessive data owner will comprise that user's the behavior of browsing, clickthrough can not directly express the action of user preferences like this.Recessive data have its certain superiority: at first recessive data do not need the extra certain operations process of user, better user experience; Recessive data are because its larger data fabrication of can effectively avoiding is brushed the deceptive practices that grade in addition.

Fig. 1 shows content association is recommended in the crawl that utilizes traditional hidden customer behavioral data for the user flow process.As shown in Figure 1, when the user browses a URL webpage, namely add this and browse behavior to the user browsing behavior database, the user browsing behavior database root is browsed the information that two aspects are obtained in behavior according to this: 1. obtain other user behavior information of browsing this webpage; 2. obtain with other similar user behavior information of this user behavior, then fall to sort according to the URL of the number of users of jointly browsing to related webpage; Demonstration rank top n URL content information is recommended the user, and the user is by clicking the URL content information that can check recommendation.

According to process flow diagram shown in Figure 1, can find, the grasping means of present traditional hidden customer behavioral data just obtains the user and clicks the content of browsing a page, but because a present page often comprised a large amount of different contents, the user has often just browsed or has paid close attention to wherein very little a part of content in the process of browsing.

And because the difference of website UI design, the focal position that different web pages user pays close attention to all there are differences.And because the restriction of the mobile device screen sizes such as mobile phone, panel computer, the content of screen display is very limited, need the user to check content different in the webpage (very extensive on mobile phone is used such as the social network sites of present hot topic, microblogging website etc., as to have comprised the different content that a lot of users produce in the page) by the mode of rolling roller bearing or slip.But be different so browse the content that the different user of the same URL page has very big the subject of knowledge and the object of knowledge to browse and pay close attention to (page part that the roll/slide page stops), and said method and can't distinguish this situation.Therefore, this crawl of only clicking the content of pages browse by the user can cause the reduction of recommendation results accuracy inevitably for according to the mode of determining content recommendation.

In addition, the result that present commending system calculates often fixedly is presented at a position of the page, the user is when rolling view page diverse location, the position of recommendation results and content can not produce according to the variation of user's focus the variation that is associated in real time, cause the user or can't see the content (content recommendation roll out screen display scope) of recommendation, or the content of recommending is irrelevant with the focus of current concern.

Summary of the invention

In view of the above problems, the purpose of this invention is to provide a kind of can Real-time Obtaining user focus and recommend the real-time exhibiting method of content association and the system of related content according to the different real-time change of focus, the user experiences and the conversion ratio of website to improve.

According to an aspect of the present invention, provide a kind of content association real-time exhibiting method, having comprised:

The screen that changes according to user terminal moves the content of the current screen display page of Real-time Obtaining user terminal; Determine the focus content of the described screen display page according to the content of the current screen display page of the user terminal that obtains; Judge whether the cache information that is associated with the focus content of the described screen display page according to the focus content of the described screen display page; If have, obtain described cache information and be shown to described user terminal as the content association of recommending to described user terminal; Otherwise, obtain other customer groups of the focus content of paying close attention to the described screen display page and other customer groups similar to the behavior of described user terminal by real-time calculating, determine the common focus of other customer groups, and be shown to described user terminal with described common focus as the content association of recommending to described user terminal.

Wherein, the screen action of changing of user terminal comprises that the user changes rolling or the sliding action of screen display content of pages, and is described according to user's rolling and sliding action, obtains the content of user's current display page, specifically refers to the content that the current display screen of user can show.Because the difference of screen resolution size, the content that screen page of each user can demonstrate is not quite similar, and needs here according to ticker position, and screen resolution calculates the shown content that goes out of active user's screen.

On the other hand, the present invention also provides a kind of content association to represent in real time system, comprising:

Screen display content obtaining unit is used for moving the content of the current screen display page of Real-time Obtaining user terminal according to the screen that changes of user terminal;

Focus content determining unit, the content that is used for the current screen display page of user terminal that obtains according to described screen display content obtaining unit is determined the focus content of the described screen display page;

The related information buffer unit is used for storing the focus content of the described screen display page and user in the click on content of the described screen display page relatedly;

The cache information administrative unit is used for judging at described related information buffer unit whether the cache information that is associated with the focus content of the described screen display page is arranged according to the focus content of the described screen display page; If have, obtain the content association that described cache information and conduct are recommended to described user terminal; Otherwise, obtain other customer groups of the focus content of paying close attention to the described screen display page and other customer groups similar to the behavior of described user terminal by real-time calculating, determine the common focus of other customer groups, and with described common focus as the content association of recommending to described user terminal;

Content association represents the unit, is used for the content association of recommending to described user terminal is shown to described user terminal.

In order to realize above-mentioned and relevant purpose, one or more aspects of the present invention comprise the feature that the back will describe in detail and particularly point out in the claims.Following explanation and accompanying drawing describe some illustrative aspects of the present invention in detail.Yet, the indication of these aspects only be some modes that can use in the variety of way of principle of the present invention.In addition, the present invention is intended to comprise all these aspects and their equivalent.

Description of drawings

By the content of reference below in conjunction with the description of the drawings and claims, and along with understanding more comprehensively of the present invention, other purpose of the present invention and result will understand and easy to understand more.In the accompanying drawings:

Fig. 1 utilizes the crawl of traditional hidden customer behavioral data to recommend the schematic flow sheet of content association for the user;

Fig. 2 is the schematic flow sheet according to the real-time exhibiting method of content association of the embodiment of the invention;

Fig. 3 for according to the embodiment of the invention roll or slide under carry out the schematic flow sheet that content association represents in real time;

Fig. 4 is the specific implementation process synoptic diagram according to the related information cache module of the embodiment of the invention;

Fig. 5 is for representing in real time the logical organization synoptic diagram of system according to the content association of the embodiment of the invention.

Identical label is indicated similar or corresponding feature or function in institute's drawings attached.

Embodiment

In the following description, for purposes of illustration, for the complete understanding to one or more embodiment is provided, many details have been set forth.Yet, clearly, can not have to realize these embodiment in the situation of these details yet.In other example, one or more embodiment for convenience of description, known structure and equipment illustrate with the form of block scheme.

Below with reference to accompanying drawing specific embodiments of the invention are described in detail.

Need to prove that related user terminal comprises that PC, mobile phone, panel computer etc. can carry out the network insertion terminal of network browsing among the present invention.

Fig. 2 shows the schematic flow sheet according to the real-time exhibiting method of content association of the embodiment of the invention.

As shown in Figure 2, the real-time exhibiting method of content association provided by the invention comprises the steps:

S210: the screen that changes according to user terminal moves the content of the current screen display page of Real-time Obtaining user terminal;

S220: the focus content of determining the screen display page according to the content of the current screen display page of the user terminal that obtains;

S230: judge whether the cache information that is associated with the focus content of the screen display page according to the focus content of the screen display page; If have, then enter step S240, otherwise, enter step S250;

S240: obtain the content association that described cache information and conduct are recommended to user terminal;

S250: obtain other customer groups of the focus content of paying close attention to the screen display page and other customer groups similar to the behavior of this user terminal by real-time calculating, determine the common focus of other customer groups, and with this common focus as the content association of recommending to this user terminal;

S260: will be shown to this user terminal to the content association that this user terminal is recommended.

Wherein, the screen action of changing of user terminal comprises that the user changes rolling or the sliding action of screen display content of pages, can realize by input medias such as mouse or touch-screens.Such as, the user utilizes mouse roller or mouse to pull the operation that roller bearing is checked diverse location content in the webpage when using the PC browser, and the user uses the equipment slip page such as touch-screen to check the content etc. of diverse location in using the mobile devices such as mobile phone, panel computer.

Wherein, calculate in real time and obtain other customer groups of paying close attention to this focus content and other customer groups similar to this user behavior, thereby obtain the common focus of other customer groups; Its specific implementation mainly comprises obtaining finds in nearest-neighbors (comprising customer group and the similar customer group of behavior of paying close attention to this focus content) and the joint act from nearest-neighbors that the user may interested content.Wherein, utilize the customer group of focus content to obtain the maximally related content of follow focus content, and utilize the similar customer group of behavior to obtain the maximally related content with this user.

Particularly, as example, Fig. 3 shows under rolling or slide and carries out the schematic flow sheet that content association represents in real time.As shown in Figure 3, according to after browsing needs and opening a URL page (step S310), at first obtain the content of user terminal current display page the user, specifically refer to the content (S320) that the current display screen of user can show.Because the screen resolution of each user terminal and the difference of screen size, the lower content that can demonstrate of screen page of each user is not quite similar, and can calculate the shown content that goes out of active user's screen according to ticker position, screen resolution.

Then, in step S330, obtain the focus content that the user pays close attention to according to the current page displaying contents that obtains, specifically refer to the most possible position that focuses on of user under a page.Because the uncertainty of the focus that the diversity of the content of the page and user pay close attention to, the focus that the user pays close attention in a page is not what be evenly distributed, but becomes log-Normal to distribute.The sight line that the user browses a webpage distributes and can come match by user's click action, user's click location is identical with the focus that the user pays close attention to, so utilize user's clickstream data and click location information, can effectively simulate the user and distribute in different page sight lines, thereby utilize probability theory to find the most possible focus of paying close attention to of user.

Along with the correcting variation of webpage design, the focus distribution that the user pays close attention to might change, and utilizes the approximating method of real-time clickstream data, can obtain in real time the most accurately focus distribution.Its circular is to suppose that the focus probability that the user pays close attention to can utilize following log-Normal to distribute, and needs to utilize Multivariate Log-Normal to distribute here:

$F(X,D,\mathrm{\μ})={\left(2\mathrm{\Π}\right)}^{-\frac{p}{2}}\×{\left|D\right|}^{-\frac{1}{2}}\×{[x_1,x_2,...{,x}_p]}^{-1}$

$\×\mathrm{Exp}(-\frac{{(\ln X-\mathrm{\μ})}^{\′D^{-1}}(\ln X-\mathrm{\μ})}{2})$

Wherein, P be X dimension (in display screen, dimension be 2, X be screen point coordinate wherein the lower left corner be initial point (0,0)), D and μ are respectively the parameters of X probability distribution equation, lnX-[1nx ₁, 1nx ₂..., lnx _p].So the log-normal on screen (2 dimension space) is distributed as:

$F(x,y)={\left(2\mathrm{\Π}\right)}^{-1}\×{\left|D\right|}^{-\frac{1}{2}}\×{[x,y]}^{-1}\×\mathrm{Exp}(-\frac{{\left(\right[\ln x,\ln y]-\mathrm{\μ})}^{\′D^{-1}}\left(\right[\ln x,\ln y]-\mathrm{\μ})}{2})$

Can find from above-mentioned formula, the probability distribution graph of log-normal depends on two parameter D and μ, thus the Function Fitting of log-normal, be according to existing user's click location, estimate the parameter in the above-mentioned formula, obtain with the maximum likelihood evaluation method herein:

If existing subscriber n click coordinate collection is

Calculate respectively mean value and the variance of this set,

$\stackrel{\‾}{x}=\frac{{\mathrm{\Σ}}_1{x}_1}{n},$ $\stackrel{\‾}{y}=\frac{{\mathrm{\Σ}}_1{y}_1}{n},$

$S_{\mathrm{xx}}=\mathrm{var}\left(x\right)=\underset{1}{\mathrm{\Σ}}{(x_1-\stackrel{\‾}{x})}^2/n,S_{\mathrm{yy}}=\mathrm{var}\left(y\right)=\underset{1}{\mathrm{\Σ}}{(y_1-\stackrel{\‾}{y})}^2/n$

$S_{\mathrm{xy}}=S_{\mathrm{yx}}=\mathrm{var}(x,y)=\underset{1}{\mathrm{\Σ}}(x_1-\stackrel{\‾}{x})(y_1-\stackrel{\‾}{y})/n$

Then parameter D can utilize formula estimation with μ and get:

$\mathrm{\μ}=[{\mathrm{\μ}}_x,{\mathrm{\μ}}_y]=[\exp (\stackrel{\‾}{x}+\frac{{S_{\mathrm{xx}}}^2}{2}),\exp (\stackrel{\‾}{y}+\frac{{S_{\mathrm{yy}}}^2}{2})]$

$D=\left[\begin{array}{cc}{D}_{\mathrm{xx}}& D_{\mathrm{xy}}\\ D_{\mathrm{xy}}& D_{\mathrm{yy}}\end{array}\right]$

$=\left[\begin{array}{cc}\exp (2\stackrel{\‾}{x}+\frac{{S_{\mathrm{xx}}}^2}{2})(\exp \left({S_{\mathrm{xx}}}^2\right)-1)& \frac{\exp [(\stackrel{\‾}{x}+\stackrel{\‾}{y})+({S_{\mathrm{xx}}}^2+{S_{\mathrm{yy}}}^2)}{2}(\exp \left({S_{\mathrm{xy}}}^2\right)-1)\\ \frac{\exp [(\stackrel{\‾}{x}+\stackrel{\‾}{y})+({S_{\mathrm{xx}}}^2+{S_{\mathrm{yy}}}^2)}{2}(\exp \left({S_{\mathrm{xy}}}^2\right)-1)& \exp (2\stackrel{\‾}{y}+\frac{{S_{\mathrm{yy}}}^2}{2})(\exp \left({S_{\mathrm{yy}}}^2\right)-1)\end{array}\right]$

Utilize the log-Nomal function of above-mentioned match, just can obtain the focus probability distribution that user's screen may be paid close attention to, the position of maximum probability is exactly the position that user's most probable is paid close attention to.

After the focus content of having obtained user's concern, judge further in the related buffer memory of focus information whether the cache information (step S340) that is associated with this focus content is arranged; If have, then from the related buffer memory of focus information, obtain cache information (step S351); If no, calculate in real time and obtain other customer groups (S352) of paying close attention to this focus content and other customer groups (S353) similar to this user behavior, thereby obtain the common focus of other customer groups.

Wherein, judge whether the cache information that is associated with this focus content in the related buffer memory of focus information, its Main Function is in order to improve the acquisition speed of content recommendation.Because rolling or slip correlation recommendation are very high for real-time requirement, if can not in time pushing, the content association of recommending represents to the user, the user probably rolls the screen display content or slides into other positions of the page, cause the content of the content of recommending and current demonstration to exist time-delay or user can't view the content of recommendation, in order to address this problem, adopt the related information cache module to improve the acquisition speed of recommendation information.

S360: inquire about according to above-described storage organization, inquire the historical content association of clicking according to focus information, and fall to sort according to number of clicks;

S370: the recommendation index of the content association of finding according to comparison method carries out sequencing display, show the content association of recommending, its particular content carries out sequencing display (recommend the higher ordering of index more forward) for the recommendation index of the content association found according to comparison method, the content association that shows simultaneously can not produce change in location along with user's rolling slide, and its content can be presented at user's current page all the time;

S380: its focus related content that estimates according to system recommendation, the user finds that own interested content recommendation and clickthrough check its detail;

S390: judge the operation of user's slides/rolls, comprise that specifically the user utilizes mouse roller or mouse to pull the operation that roller bearing is checked diverse location content in the webpage when using the PC browser, also comprise the content that the user uses the equipment slip page such as touch-screen to check diverse location in using the mobile devices such as mobile phone, panel computer.If detect slides/rolls operation then repeat above step, otherwise continue to carry out active user slides/rolls operation detection.

Use the focus content of related information cache module storage user concern and judge whether the focus content has the related information of buffer memory; If have, obtain cache information; If no, calculate in real time and obtain other customer groups of paying close attention to this focus content and other customer groups similar to this user behavior, thereby obtain the common focus of other customer groups.Adopt the related information cache module to improve the acquisition speed of recommendation information among the present invention, the specific implementation step of this module is divided into mainly that cache information writes and the cache information inquiry.Fig. 4 shows the specific implementation step according to the related information cache module of the embodiment of the invention.

As shown in Figure 4, the process according to the related information buffer memory of the related information cache module of the embodiment of the invention and related information inquiry is as follows:

Step S411-S412: write for cache information, mainly click behavior according to the user for the history of content recommendation, its concrete steps can be described as, when a page of user's rolling view, related content is recommended out according to the focus that the user who estimates may pay close attention to by system, the user finds that own interested content recommendation can check its detail by clickthrough, at this moment (one record comprises the focus content as Key the content of clicking to be carried out association store with the focus content of the current concern of user, click on content is as Value), add up simultaneously the historical number of clicks (number of clicks that comprises other users) of this click on content, after number of clicks surpasses certain threshold values (for example above 10 times), just (the focus content is as Key this association store content, click on content with and number of clicks as Value) put in the buffer memory in order to the fast query of content recommendation and obtain, simultaneously in order to guarantee that cache size can unrestrictedly not enlarge, after buffer memory capacity surpassed certain threshold values, the record that number of clicks is minimum can remove from buffer memory;

Step S420: the cache information inquiry, mainly inquire about according to above-described storage organization, inquire the historical content association of clicking according to focus information, and fall to sort according to number of clicks.Corresponding with the real-time exhibiting method of above-mentioned content association, the present invention also provides a kind of content association to represent in real time system.Fig. 5 shows the logical organization synoptic diagram that represents in real time system according to the content association of the embodiment of the invention.

As shown in Figure 5, content association provided by the invention represents in real time system 500 and comprises that screen display content obtaining unit 510, focus content determining unit 520, related information buffer unit 530, cache information administrative unit 540 and content association represent unit 550.

Wherein, screen display content obtaining unit 510 is used for moving the content of the current screen display page of Real-time Obtaining user terminal according to the screen that changes of user terminal; Focus content determining unit 520 is determined the focus content of the screen display page for the content of the current screen display page of user terminal that obtains according to screen display content obtaining unit 510; Related information buffer unit 530 is used for storing the focus content of the screen display page and user in the click on content of the screen display page relatedly; Cache information administrative unit 540 is used for judging at related information buffer unit 530 whether the cache information that is associated with the focus content of the above-mentioned screen display page is arranged according to the focus content of the screen display page; If have, obtain the content association that this cache information and conduct are recommended to user terminal; Otherwise, obtain other customer groups of the focus content of paying close attention to the screen display page and other customer groups similar to the behavior of user terminal by real-time calculating, determine the common focus of other customer groups, and with common focus as the content association of recommending to described user terminal.At last, represent unit 550 by content association the content association of recommending to user terminal is shown to user terminal.

Wherein, all the time can be to the user as seen in order to make representing of content recommendation, the content association that content association represents unit 550 demonstrations can not produce change in location along with user's rolling slide, and its content can be presented at user's current page all the time.

As above in the mode of example the real-time exhibiting method of content association according to the present invention and system are described with reference to accompanying drawing.But, it will be appreciated by those skilled in the art that the real-time exhibiting method of the content association that proposes for the invention described above and system, can also make various improvement on the basis that does not break away from content of the present invention.Therefore, protection scope of the present invention should be determined by the content of appending claims.

Claims (10)

1. real-time exhibiting method of content association comprises:

The screen that changes according to user terminal moves the content of the current screen display page of Real-time Obtaining user terminal;

Determine the focus content of the described screen display page according to the content of the current screen display page of the user terminal that obtains;

Judge whether the cache information that is associated with the focus content of the described screen display page according to the focus content of the described screen display page; If have, obtain described cache information and be shown to described user terminal as the content association of recommending to described user terminal; Otherwise,

Obtain other customer groups of the focus content of paying close attention to the described screen display page and other customer groups similar to the behavior of described user terminal by real-time calculating, determine the common focus of other customer groups, and be shown to described user terminal with described common focus as the content association of recommending to described user terminal.

2. the real-time exhibiting method of content association as claimed in claim 1, wherein, the screen action of changing of user terminal comprises that the user changes rolling or the sliding action of screen display content of pages.

3. the real-time exhibiting method of content association as claimed in claim 2, wherein, the process of obtaining the content of user's current display page according to rolling or the sliding action of user terminal comprises:

Determine the content of the current screen display page of described user terminal according to the screen resolution of ticker position and described user terminal in the described screen display page.

4. the real-time exhibiting method of content association as claimed in claim 1, wherein, described content according to the current screen display page of the user terminal that obtains determines that the process of the focus content of the described screen display page comprises:

Utilize clickstream data and the click location information of user terminal, go out the user according to the log-Normal Function Fitting of match and distribute in the sight line of the current screen display page;

Utilize probability theory with the position of the sight line distribution probability maximum focus content as the determined described screen display page.

5. the real-time exhibiting method of content association as claimed in claim 4 wherein, goes out the user according to the log-Normal Function Fitting of match and comprises in the process that the sight line of the current screen display page distributes:

Utilize Multivariate Log-Normal to distribute:

Wherein, P is the dimension of X, and D and μ are respectively the parameters of X probability distribution equation, and in the display screen of user terminal, dimension is that 2, X is the coordinate of screen point, and wherein the lower left corner is initial point (0,0);

The log-normal that obtains on screen 2 dimension spaces is distributed as:

6. real time correlation content display method as claimed in claim 1 wherein, after determining the focus content of the described screen display page according to the content of the current screen display page of the user terminal that obtains, also comprises:

Deposit the focus content of the described screen display page in buffer memory with the user in the click on content of the described screen display page relatedly.

7. real time correlation content display method as claimed in claim 6, wherein, obtain described cache information and comprise as the process that the content association of recommending to described user terminal is shown to described user terminal:

Obtain described cache information, described cache information comprises that the focus content of the screen display page of association store and user are in the click on content of the described screen display page;

The number of clicks of the content association of the history that inquires being clicked according to described focus content sorts;

Focus content with the predetermined number that number of clicks is the highest in the described ordering is shown to described user terminal as the content association of recommending to described user terminal.

8. the real-time exhibiting method of content association as claimed in claim 1, wherein, calculate in real time and obtain other customer groups of the focus content of paying close attention to the described screen display page and other customer groups similar to the behavior of described user terminal, determine that the process of the common focus of other customer groups comprises:

Obtain and find the interested content of user in nearest-neighbors and the joint act from nearest-neighbors, described nearest-neighbors comprises customer group and the similar customer group of behavior of paying close attention to described focus content;

According to the user terminal quantity of jointly browsing described focus content is sorted; With the focus content of the highest predetermined number of the user terminal quantity jointly browsed in the described ordering as described common focus; Wherein,

Utilize the customer group of focus content to obtain and the maximally related content of described focus content, utilize the similar customer group of behavior to obtain the maximally related content with this user.

9. the real-time exhibiting method of content association as claimed in claim 1, wherein, the described content association that is shown to the recommendation of described user terminal is presented on the current screen display page of user terminal all the time.

10. a content association represents system in real time, comprising:

Screen display content obtaining unit is used for moving the content of the current screen display page of Real-time Obtaining user terminal according to the screen that changes of user terminal;

Focus content determining unit, the content that is used for the current screen display page of user terminal that obtains according to described screen display content obtaining unit is determined the focus content of the described screen display page;

The related information buffer unit is used for storing the focus content of the described screen display page and user in the click on content of the described screen display page relatedly;

The cache information administrative unit is used for judging at described related information buffer unit whether the cache information that is associated with the focus content of the described screen display page is arranged according to the focus content of the described screen display page; If have, obtain the content association that described cache information and conduct are recommended to described user terminal; Otherwise, obtain other customer groups of the focus content of paying close attention to the described screen display page and other customer groups similar to the behavior of described user terminal by real-time calculating, determine the common focus of other customer groups, and with described common focus as the content association of recommending to described user terminal;

Content association represents the unit, is used for the content association of recommending to described user terminal is shown to described user terminal.

Images