CN103108049B - A kind ofly provide the method for personal page for mobile phone users - Google Patents

Abstract

For mobile phone users provides a method for personal page, include: step one, build several different content of pages; Step 2, from the identity characteristic of mobile phone users, obtain several characteristic indexs, and based on the membership function in fuzzy mathematics, calculate the degree of membership that user belongs to the basic, normal, high end subscriber group of each characteristic index respectively, then according to each page for each characteristic index basic, normal, high end subscriber group stress accounting, finally calculate the matching degree of each characteristic index of user and each page; Step 3, according to the matching degree of each characteristic index of user and each page and weighted value corresponding to each characteristic index of user, calculate the matching degree of user and each page, and provide personalized content of pages to user accordingly.The invention belongs to net application technology field, according to the identity characteristic of mobile phone users for user provides personal page, can also be applied to the portal website of books reading further.

Description

A kind ofly provide the method for personal page for mobile phone users

Technical field

The present invention relates to a kind ofly provides the method for personal page for mobile phone users, belongs to net application technology field.

Background technology

Day by day increase severely instantly in the network information, how accurate assurance mobile phone users demand, the information meeting user preferences is pushed in face of user with air line distance seems particularly important fast.How for mobile phone users provides personal page, it is also proposed the solution that some are relevant at present:

Patent application CN 201010179183.X (application title: a kind of pushing personalized pages to mobile terminal method and apparatus, the applying date: 2010-05-19, applicant: Shenzhou Wuju Technology Co., Ltd.) disclose a kind of pushing personalized pages to mobile terminal method and apparatus, described method comprises user behavior statistical analysis process and page push process, described page push process comprises: step 21: user clicks the page, and mobile terminal is connected to remote server; Step 22: obtain user access header, analyze wherein with label; Step 23: obtain one or more correlation tag the highest with the described label degree of association; Step 24: according to the keyword in above-mentioned label and correlation tag, finds corresponding resource in Resource Server, searches out the object content of user's request and personalized push content and forms the final page and feed back to user.This technical scheme is mainly used in pushing personalized Webpage according to the use preference (i.e. Visitor Logs) of user, because user has age, individual character, the inferior different identity characteristic of stratum of intellectual, its hobby also varies with the difference of identity, and this technical scheme can not provide personal page according to the identity characteristic of user (such as the moon communication fee, note amount, GPRS flow or age etc. can the characteristic indexs of characterizing consumer identity characteristic).

Therefore, how according to the identity characteristic of mobile phone users, for user provides personal page, a unsolved technical problem is still.

Summary of the invention

In view of this, the object of this invention is to provide a kind ofly provides the method for personal page for mobile phone users, can according in the identity characteristic of mobile phone users for user provides personal page.

In order to achieve the above object, the invention provides a kind ofly provides the method for personal page for mobile phone users, includes:

Step one, build several different content of pages;

Step 2, from the identity characteristic of mobile phone users, obtain several characteristic indexs, and based on the membership function in fuzzy mathematics, calculate the degree of membership that user belongs to the basic, normal, high end subscriber group of each characteristic index respectively, then according to each page for each characteristic index basic, normal, high end subscriber group stress accounting, finally calculate the matching degree of each characteristic index of user and each page;

Step 3, according to the matching degree of each characteristic index of user and each page and weighted value corresponding to each characteristic index of user, calculate the matching degree of user and each page, and provide personalized content of pages to user accordingly,

In described step 2, user belongs to i-th characteristic index tz _ithe degree of membership ptz of low-end users _i_ low (x _i) computing formula as follows:

${\mathrm{ptz}}_i\_\mathrm{low}\left(x_i\right)=\left\{\begin{array}{cc}\frac{x_i}{a},& 0<x_i\&le;a\\ 1,& a<x_i\&le;b\\ \frac{c-x_i}{c-b},& b<x_i\&le;c\\ 0,& x_i>c\end{array}\right.;$

User belongs to i-th characteristic index tz _ithe degree of membership ptz of end users group _i_ mid (x _i) computing formula as follows:

${\mathrm{ptz}}_i\_\mathrm{mid}\left(x_i\right)=\left\{\begin{array}{cc}0,& 0<x_i\&le;b\\ \frac{x_i-b}{c-b},& b<x_i\&le;c\\ 1,& c<x_i\&le;d\\ \frac{e-x_i}{e-d},& d<x_i\&le;e\\ 0,& x_i>e\end{array}\right.;$

User belongs to i-th characteristic index tz _ithe degree of membership ptz of high-end user group _i_ high (x _i) computing formula as follows:

${\mathrm{ptz}}_i\_\mathrm{high}\left(x_i\right)=\left\{\begin{array}{cc}0,& 0<x_i\&le;d\\ \frac{x_i-d}{e-d},& d<x_i\&le;e\\ 1,& e<x_i\&le;f\\ \frac{g-x_i}{g-f},& f<x_i\&le;g\\ 0,& x_i>g\end{array}\right.;$

Wherein, x _ithat user corresponds to i-th characteristic index tz _ivalue, interval parameter a, b, c, d, e, f, g be greater than 0 real number.

Compared with prior art, the invention has the beneficial effects as follows: the division due to customer group has ambiguity and uncertainty, the present invention adopts membership function in fuzzy mathematics as the method distinguishing customer group, moon communication fee is obtained from user identity feature, note amount, multiple characteristic indexs such as GPRS flow and age, and calculate user and belong to the low of each characteristic index, in, the degree of membership of high-end user group, that is, the level index that user is under the jurisdiction of certain page is weighed by degree of membership, the size of the corresponding probability of size of degree of membership, namely the larger probability of degree of membership is larger, thus can according to the identity characteristic of user for user provides the personal page conformed to most with its hobby, achieve the guiding directly perceived to user preferences content, for the portal website of books reading, the present invention also further by user's past to book content on the page, recommend the visit record of position link or activity to analyze the reading behavior preference of user, and the matching degree of user and the page is adjusted according to the reading behavior preference of user, thus can in user identity feature base, the read data of further integration user, comprehensively defines the books reading page meeting user preferences most.

Accompanying drawing explanation

Fig. 1 is that the present invention is a kind of provides the method flow diagram of personal page for mobile phone users.

Fig. 2 is in Fig. 1 step 3, when being applied to books reading portal website, according to the particular flow sheet that the book content of the page matching degree to user and the page adjusts.

Fig. 3 is in Fig. 1 step 3, when being applied to books reading portal website, according to the particular flow sheet that the matching degree of the recommendation position of page link to user and the page adjusts.

Fig. 4 is in Fig. 1 step 3, when being applied to books reading portal website, according to the particular flow sheet that the activity of the page matching degree to user and the page adjusts.

Embodiment

For making the object, technical solutions and advantages of the present invention clearly, below in conjunction with accompanying drawing, the present invention is described in further detail.

As shown in Figure 1, the present invention is a kind of provides the method for personal page for mobile phone users, includes:

Step one, build several different content of pages;

Step 2, from the identity characteristic of mobile phone users, obtain several characteristic indexs, and based on the membership function in fuzzy mathematics, calculate the degree of membership that user belongs to the basic, normal, high end subscriber group of each characteristic index respectively, then according to each page for each characteristic index basic, normal, high end subscriber group stress accounting, finally calculate the matching degree of each characteristic index of user and each page;

Step 3, according to the matching degree of each characteristic index of user and each page and weighted value corresponding to each characteristic index of user, calculate the matching degree of user and each page, and provide personalized content of pages to user accordingly.

In described step one, multiple content of pages can be built according to actual business requirement, such as 4 pages.

In described step 2, because user identity can be middle school student's (comprising junior-senior high school raw), university student, alien workers, middle and high end business people (such as white collar and some social elites) etc., these different user identity have larger difference in age composition, the level of consumption, online demand etc., and the characteristic index therefore obtained from the identity characteristic of mobile phone users can be a month communication fee, note amount, GPRS flow or age etc.

Division due to customer group has ambiguity and uncertainty, as white collar and alien workers, white collar and high-end business people do not have clear and definite distinguishing limit; Secondly user identity feature and the page do not have clear and definite belonging relation yet, and as university student not must see the content of struggling and pursuing a goal with determination, high-end business people also not only sees the content through pipe history.Therefore, the present invention adopts membership function in fuzzy mathematics as the method distinguishing customer group, and that is, weigh by degree of membership the level index that user is under the jurisdiction of certain page, the size of the corresponding probability of size of degree of membership, namely the larger probability of degree of membership is larger.

In step 2, user belongs to i-th characteristic index tz _ithe degree of membership ptz of low-end users _i_ low (x _i) computing formula as follows:

${\mathrm{ptz}}_i\_\mathrm{low}\left(x_i\right)=\left\{\begin{array}{cc}\frac{x_i}{a},& 0<x_i\&le;a\\ 1,& a<x_i\&le;b\\ \frac{c-x_i}{c-b},& b<x_i\&le;c\\ 0,& x_i>c\end{array}\right.;$

User belongs to i-th characteristic index tz _ithe degree of membership ptz of end users group _i_ mid (x _i) computing formula as follows:

${\mathrm{ptz}}_i\_\mathrm{mid}\left(x_i\right)=\left\{\begin{array}{cc}0,& 0<x_i\&le;b\\ \frac{x_i-b}{c-b},& b<x_i\&le;c\\ 1,& c<x_i\&le;d\\ \frac{e-x_i}{e-d},& d<x_i\&le;e\\ 0,& x_i>e\end{array}\right.;$

User belongs to i-th characteristic index tz _ithe degree of membership ptz of high-end user group _i_ high (x _i) computing formula as follows:

${\mathrm{ptz}}_i\_\mathrm{high}\left(x_i\right)=\left\{\begin{array}{cc}0,& 0<x_i\&le;d\\ \frac{x_i-d}{e-d},& d<x_i\&le;e\\ 1,& e<x_i\&le;f\\ \frac{g-x_i}{g-f},& f<x_i\&le;g\\ 0,& x_i>g\end{array}\right.;$

Wherein, x _ithat user corresponds to i-th characteristic index tz _ivalue, interval parameter a, b, c, d, e, f, g be greater than 0 real number, its value sets according to business actual conditions.

Such as when characteristic index is month communication fee, the value of described interval parameter can be set to: a=25, b=40, c=45, d=60, e=80, f=125, g=800, x ₁the moon communication fee value of user, namely user to belong to the computing formula of the degree of membership of the low-end users of moon communication fee as follows: ${\mathrm{ptz}}_1\_\mathrm{low}\left(x_1\right)=\left\{\begin{array}{c}\frac{x_1}{25},0<x_1\&le;25\\ \mathrm{1,25}<x_1\&le;40\\ 9-\frac{x_1}{5},40<x_1\&le;45\\ 0,x_1>45\end{array}\right.;$ The computing formula that user belongs to the degree of membership of the end users group of moon communication fee is as follows: ${\mathrm{ptz}}_1\_\mathrm{mid}\left(x_1\right)=\left\{\begin{array}{c}\mathrm{0,0}<x_1\&le;40\\ \frac{x_1}{5}-\mathrm{8,40}<x_1\&le;45\\ \mathrm{1,45}<x_1\&le;60\\ 4-\frac{x_1}{20},60<x_1\&le;80\\ 0,x_1>80\end{array}\right.;$ The computing formula that user belongs to the degree of membership of the high-end user group of moon communication fee is as follows: ${\mathrm{ptz}}_1\_\mathrm{high}\left(x_1\right)=\left\{\begin{array}{c}\mathrm{0,0}<x_1\&le;60\\ \frac{x_1}{20}-\mathrm{3,60}<x_1\&le;80\\ \mathrm{1,80}<x_1\&le;125\\ \frac{800-x_1}{675},125<x_1\&le;800\\ 0,x_1>800\end{array}\right..$

Equally, when characteristic index is note amount, the value of described interval parameter can be set to: a=5, b=10, c=15, d=30, e=60, f=140, g=800; When characteristic index is GPRS flow, the value of described interval parameter can be set to: a=5, b=10, c=18, d=30, e=35, f=55, g=400; When characteristic index is the age, the value of described interval parameter can be set to: a=5, b=10, c=15, d=30, e=60, f=140, g=80.

For ensureing the demand of page setup As soon as possible Promising Policy most users differentiation, namely constructed multiple pages will give priority to and be inclined in the high-, middle-and low-end customer group of each characteristic index, the such as page one lays particular emphasis on the customer group of high moon communication fee, low note amount, high GPRS flow and middle high age bracket, such one side can promote the value identity of user, also can consider many psychology that becomes of fellow users on the other hand.Therefore, can according to actual business requirement, adopt analysis expert method pre-set each page for each characteristic index basic, normal, high end subscriber group stress accounting, and each page is 1 for the accounting sum that stresses of the basic, normal, high end subscriber group of each characteristic index, that is, p _j_ tz _i_ l+p _j_ tz _i_ m+p _j_ tz _i_ h=1, wherein p _j_ tz _i_ l, p _j_ tz _i_ m, p _j_ tz _i_ h is that page j is for characteristic index tz respectively _ibasic, normal, high end subscriber group stress accounting.Such as shown in following table 1, build 4 pages: the page one, the page two, the page three and the page four, and from user identity feature, obtain characteristic index be: the moon communication fee, note amount, GPRS flow and age, then according to actual business requirement, to pre-set in table 14 pages respectively for the moon communication fee, note amount, GPRS flow and the basic, normal, high end subscriber group at age stress accounting, because the page one lays particular emphasis on the customer group of high moon communication fee, then p ₁_ tz ₁_ l=0.2, p ₁_ tz ₁_ m=0.2, p ₁_ tz ₁_ h=0.6.

In described step 2, user i-th characteristic index tz _iwith the matching degree p of page j _jtz _i(x _i) computing formula as follows:

P _jtz _i(x _i)=ptz _i_ low (x _i) * p _j_ tz _i_ l+ptz _i_ mid (x _i) * p _j_ tz _i_ m+ptz _i_ high (x _i) * p _j_ tz _i_ h, wherein, x _ithat user corresponds to i-th characteristic index tz _ivalue, ptz _i_ low (x _i), ptz _i_ mid (x _i), ptz _i_ high (x _i) be that user belongs to i-th characteristic index tz respectively _ithe degree of membership of basic, normal, high end subscriber group, p _j_ tz _i_ l, p _j_ tz _i_ m, p _j_ tz _i_ h is that page j is for characteristic index tz respectively _ibasic, normal, high end subscriber group stress accounting.

In described step 3, the computing formula of the matching degree of user and page j is as follows:

wherein, ω _iuser i-th characteristic index tz _icorresponding weighted value, n is the characteristic index sum obtained from user identity feature.From the matching degree of user and all pages, select maximum, and the content of pages corresponding to matching degree maximum is supplied to mobile phone users, thus realize personalized Webpage.

It is worth mentioning that, the present invention can also be applied in the portal website of books reading further.For the portal website of books reading, the Main Differences of different content of pages comprises book content, recommends position link and activity, wherein:

(1), book content: content tab storehouse can be set up according to books theme, plot, and by several content tabs in content tab storehouse, the book content of user-accessible on the page be identified;

(2), position link is recommended: user by clicking the recommendation position link of the page, can jump to first level pages;

(3), movable: user initiatively can participate in the activity of multi-form content on the page.

When the present invention is applied in the portal website of books reading, the present invention can also on the identity characteristic basis of user, further by analyze user's past to the book content on the page, recommend position link or the visit record of activity to analyze the reading behavior preference of user, and adjust the matching degree of user and the page according to the reading behavior preference of user.

As shown in Figure 2, in described step 3, when being applied to books reading portal website, the present invention can also adjust according to the matching degree of the book content of the page to user and page j, includes further:

Steps A 1, the unduplicated content tab comprised by book contents all on page j form page j content tab collection, and the book information that inquiring user was accessed within a period of time (as nearly 3 months), then the frequency occurred in the content tab concentrating each content tab to comprise at all visited book content of user according to page j content tab, calculate user-content tab weight that page j content tab concentrates each content tab one by one, namely wherein tr _μuser-content tab weight that page j content tab concentrates μ content tab, t _μit is the occurrence number of μ content tab in the content tab comprised at all visited book content of user, T is the appearance total degree of all the elements label that all visited book content of user comprises, the content tab that wherein book content comprises chooses from content tab storehouse, and the keyword for identifying book content;

Steps A 2, adjustment user and the matching degree of page j: wherein, because user identity characteristic sum reading behavior preference is different to the influence degree of the page, therefore α is the weight of user identity feature, 1-α is the weight of user's reading behavior preference, λ is the importance measures weight of book content for page Division identification, nj is the content tab sum that page j content tab is concentrated, tr _μuser-content tab weight that page j content tab concentrates μ content tab, R _{j μ}be the weight that page j content tab concentrates μ content tab, because the book content of the different page may exist intersection, for ensureing that the page is distinguished, the weight of the same book content of the different page may be different, and namely the weight of same content tab may be different, R _{j μ}value set according to practical business situation.

As shown in Figure 3, in described step 3, when being applied to books reading portal website, the present invention can also adjust the matching degree of user and page j according to the recommendation position link of the page, includes further:

Step B1, inquiring user within a period of time (as nearly 3 months) to access (the namely click) information of recommending position link, and the frequency occurred in user's all visited recommendation position link is linked at according to each recommendation position of page j, calculate the weight of each recommendation position link of page j one by one, namely wherein lr _νν the weight of recommending position to link of page j, l _νbe ν access times of recommending position to link in user's all visited recommendation position link, L is the access total degree that user links all recommendation positions;

Step B2, adjustment user and the matching degree of page j: wherein, α is the weight of user identity feature, and 1-α is the weight of user's reading behavior preference, and β recommends position link for the importance measures weight of page Division identification, m _jthe recommendation position links total number of page j, lr _νν the weight of recommending position to link of page j.

In like manner, as shown in Figure 4, in described step 3, when being applied to books reading portal website, the present invention can also adjust according to the matching degree of the activity of the page to user and page j, includes further:

The action message of step C1, inquiring user active participate within a period of time (as nearly 3 months), and according to the frequency that each activity of page j occurs in all activities participated in of user, calculate the weight of each activity of page j one by one, namely wherein ar _ηthe weight of η the activity of page j, a _ηbe the participation number of times of η activity in all activities participated in of user, A is the participation total degree of user to all activities;

Step C2, adjustment user and the matching degree of page j: wherein, α is the weight of user identity feature, and 1-α is the weight of user's reading behavior preference, and γ is the movable importance measures weight for page Division identification, d _jthe activity sum of page j, ar _ηit is the weight of η the activity of page j.

In order to the technique effect of clearer explanation the inventive method, applicant is extracted the BOSS identity data of certain customers and the reading access interactive data of nearly 3 months.After application the inventive method, calculate the matching degree that user and 4 build the page respectively, as shown in the table, from the matching degree of each user and 4 pages, select maximum, and provide the content of pages corresponding to described maximum to user.

User	The page one	The page two	The page three	The page four	The page is provided
						User1	0.1600	0.1867	0.6000	0.0533	The page three
User2	0.3014	0.1791	0.5129	0.0067	The page three
						User3	0.6177	0.2727	0.1096	0.0000	The page one

User4	0.3902	0.4061	0.1852	0.0185	The page two
						User5	0.2087	0.3740	0.2957	0.1216	The page two
User6	0.3420	0.1961	0.4508	0.0110	The page three
						User7	0.4920	0.2320	0.2760	0.0000	The page one
User8	0.2653	0.3800	0.0295	0.3253	The page two
						User9	0.1600	0.1867	0.5834	0.0700	The page three
User10	0.1200	0.1600	0.7200	0.0000	The page three
						User11	0.1226	0.2200	0.1574	0.4999	The page four
User12	0.0147	0.2619	0.0073	0.7161	The page four
						User13	0.2737	0.3002	0.2303	0.1958	The page two
User14	0.4898	0.0327	0.4775	0.0001	The page one

From service data, the present invention can solve the deficiencies in the prior art, achieves goal of the invention and good technique effect.

Above are only preferred embodiment of the present invention, not in order to limit the present invention, within the spirit and principles in the present invention all, any amendment made, equivalent replacement, improvement etc., all should be included within the scope of protection of the invention.

Claims (10)

1., for mobile phone users provides a method for personal page, it is characterized in that, include:

Step one, build several different content of pages;

Step 2, from the identity characteristic of mobile phone users, obtain several characteristic indexs, and based on the membership function in fuzzy mathematics, calculate the degree of membership that user belongs to the basic, normal, high end subscriber group of each characteristic index respectively, then according to each page for each characteristic index basic, normal, high end subscriber group stress accounting, finally calculate the matching degree of each characteristic index of user and each page;

Step 3, according to the matching degree of each characteristic index of user and each page and weighted value corresponding to each characteristic index of user, calculate the matching degree of user and each page, and provide personalized content of pages to user accordingly,

In described step 2, user belongs to i-th characteristic index tz _ithe degree of membership ptz of low-end users _i_ low (x _i) computing formula as follows:

${\mathrm{ptz}}_i\_\mathrm{low}\left(x_i\right)=\left\{\begin{array}{cc}\frac{x_i}{a},& 0<x_i\&le;a\\ 1,& a<x_i\&le;b\\ \frac{c-x_i}{c-b},& b<x_i\&le;c\\ 0,& x_i>c\end{array}\right.;$

User belongs to i-th characteristic index tz _ithe degree of membership ptz of end users group _i_ mid (x _i) computing formula as follows:

${\mathrm{ptz}}_i\_\mathrm{mid}\left(x_i\right)=\left\{\begin{array}{cc}0,& 0<x_i\&le;b\\ \frac{x_i-b}{c-b},& b<x_i\&le;c\\ 1,& c<x_i\&le;d\\ \frac{e-x_i}{e-d},& d<x_i\&le;e\\ 0,& x_i>e\end{array}\right.;$

User belongs to i-th characteristic index tz _ithe degree of membership ptz of high-end user group _i_ high (x _i) computing formula as follows:

${\mathrm{ptz}}_i\_\mathrm{high}\left(x_i\right)=\left\{\begin{array}{cc}0,& 0<x_i\&le;d\\ \frac{x_i-d}{e-d},& d<x_i\&le;e\\ 1,& e<x_i\&le;f\\ \frac{g-x_i}{g-f},& f<x_i\&le;g\\ 0,& x_i>g\end{array}\right.;$

Wherein, x _ithat user corresponds to i-th characteristic index tz _ivalue, interval parameter a, b, c, d, e, f, g be greater than 0 real number.

2. method according to claim 1, is characterized in that, in described step 2, the characteristic index obtained from the identity characteristic of mobile phone users includes but not limited to a moon communication fee, note amount, GPRS flow or age.

3. method according to claim 1, is characterized in that, when characteristic index is month communication fee, the value of described interval parameter is set to: a=25, b=40, c=45, d=60, e=80, f=125, g=800; When characteristic index is note amount, the value of described interval parameter is set to: a=5, b=10, c=15, d=30, e=60, f=140, g=800; When characteristic index is GPRS flow, the value of described interval parameter is set to: a=5, b=10, c=18, d=30, e=35, f=55, g=400; When characteristic index is the age, the value of described interval parameter is set to: a=5, b=10, c=15, d=30, e=60, f=140, g=80.

4. method according to claim 1, it is characterized in that, in described step 2, pre-set each page for each characteristic index basic, normal, high end subscriber group stress accounting, and each page is 1 for the accounting sum that stresses of the basic, normal, high end subscriber group of each characteristic index.

5. method according to claim 1, is characterized in that, in described step 2, and user i-th characteristic index tz _iwith the matching degree p of page j _jtz _i(x _i) computing formula as follows:

P _jtz _i(x _i)=ptz _i_ low (x _i) * p _j_ tz _i_ l+ptz _i_ mid (x _i) * p _j_ tz _i_ m+ptz _i_ high (x _i) * p _j_ tz _i_ h, wherein, x _ithat user corresponds to i-th characteristic index tz _ivalue, ptz _i_ low (x _i), ptz _i_ mid (x _i), ptz _i_ high (x _i) be that user belongs to i-th characteristic index tz respectively _ithe degree of membership of basic, normal, high end subscriber group, p _j_ tz _i_ l, p _j_ tz _i_ m, p _j_ tz _i_ h is that page j is for characteristic index tz respectively _ibasic, normal, high end subscriber group stress accounting.

6. method according to claim 1, is characterized in that, in described step 3, the computing formula of the matching degree of user and page j is as follows: wherein, ω _iuser i-th characteristic index tz _icorresponding weighted value, n is the characteristic index sum obtained from user identity feature, p _jtz _i(x _i) be user i-th characteristic index tz _iwith the matching degree of page j.

7. method according to claim 1, is characterized in that, described step 3 also includes:

From the matching degree of user and all pages, select maximum, and the content of pages corresponding to matching degree maximum is supplied to mobile phone users.

8. method according to claim 1, is characterized in that, when being applied to books reading portal website, the matching degree of book content to user and page j according to the page adjusts, and described step 3 also includes:

Steps A 1, the unduplicated content tab comprised by book contents all on page j form page j content tab collection, and the book information that inquiring user was accessed within a period of time, then the frequency occurred in the content tab concentrating each content tab to comprise at all visited book content of user according to page j content tab, calculates user-content tab weight that page j content tab concentrates each content tab one by one;

Steps A 2, adjustment user and the matching degree of page j: wherein, α is the weight of user identity feature, and 1-α is the weight of user's reading behavior preference, and λ is the importance measures weight of book content for page Division identification, n _jthe content tab sum that page j content tab is concentrated, tr _μuser-content tab weight that page j content tab concentrates μ content tab, R _{j μ}it is the weight that page j content tab concentrates μ content tab.

9. method according to claim 1, is characterized in that, when being applied to books reading portal website, the recommendation position link according to the page adjusts the matching degree of user and page j, and described step 3 also includes:

Step B1, inquiring user to the visit information recommending position link, and are linked at the frequency occurred in the link of user all visited recommendation position within a period of time according to each recommendation position of page j, calculate the weight of each recommendation position link of page j one by one;

Step B2, adjustment user and the matching degree of page j: wherein, α is the weight of user identity feature, and 1-α is the weight of user's reading behavior preference, and β recommends position link for the importance measures weight of page Division identification, m _jthe recommendation position links total number of page j, lr _νν the weight of recommending position to link of page j.

10. method according to claim 1, is characterized in that, when being applied to books reading portal website, the matching degree of activity to user and page j according to the page adjusts, and described step 3 also includes:

The action message of step C1, inquiring user active participate within a period of time, and according to the frequency that each activity of page j occurs in all activities participated in of user, calculate the weight of each activity of page j one by one;

Step C2, adjustment user and the matching degree of page j: wherein, α is the weight of user identity feature, and 1-α is the weight of user's reading behavior preference, and γ is the movable importance measures weight for page Division identification, d _jthe activity sum of page j, ar _ηit is the weight of η the activity of page j.