CN103399918B - A kind of method improving the searched rate in website - Google Patents

Abstract

A kind of method improving the searched rate in website of the present invention, belong to the Internet and searching engine field, the present invention affects factor and the pattern of search engine by collecting, linear iterative algorithm is utilized to be iterated obtaining weights, thus website ranking is drawn a mark that can quantify, and for the value of each influence factor with the product of weights, the reasonable proposal to web information flow can be proposed, reduce cost, improve the efficiency of website promotion.

Description

A kind of method improving the searched rate in website

Technical field

The invention belongs to the Internet and searching engine field, be specifically related to a kind of method improving the searched rate in website.

Background technology

Development along with internet economy, " with search engine as platform; by adjust webpage in search results pages ranking thus bring visit capacity " have become as a kind of marketing mode, the most so-called search engine marketing (SEM, Search Engine Marketing).SEM mainly includes business promotion optimization and search engine optimization (SEO, the Search Engine Optimization) both of which paid.The SEM paid refers generally to bid ranking mechanism, and SEO refer to promote webpage in search engine nature Search Results (non-commercial Extended Results) include quantity and sorting position and the optimization behavior done.This behavior can bring the lifting of Consumer's Experience and conversion ratio for website, the self-marketing solutions of ecological type is provided to website simultaneously, allow website position oneself at the forefront in industry, thus obtain brand income, particularly in the case of the enterprise web site enough funds of shortage carry out publicity popularization, SEO is undoubtedly a kind of low cost, instant effect, optimal website promotion method that effect is lasting, therefore suffers from the emphasis of more and more enterprise.And along with Baidu is increasingly severe about the examination & verification of bid ranking, a lot of bid rankings are all adjusted, this has attracted more website operator to be sought for by web information flow to promote ranking in a search engine.

But, in the real process that SEO optimizes, website webmaster implements to optimize the most by rule of thumb, and it is slow not only to consume manager's energy but also speed, and cost is high, reduces the creativity of enterprise self.Therefore, it is necessary to one instrument automatically analyzing website SEO mass of exploitation, auxiliary web site manager carries out SEO and optimizes.

Summary of the invention

For the deficiencies in the prior art, the present invention proposes a kind of method improving the searched rate in website, to reach to make website ranking quantify, reduces cost, specifies solution, improve the purpose of the efficiency of website promotion.

A kind of method improving the searched rate in website, comprises the following steps:

Step 1, determine the factor affecting site search engine masses, including scripted code line number in the pagefile size of tested website, the page and the ratio of total code line number, the ratio of title and the size of motion picture files in the matching degree of keyword, website in Website page with the size of animation file in the ratio of Website page file size, website with Website page size of tested website, and by the way of iteration, determine the threshold value of above-mentioned five each scores of factor respectively；

Step 1-1, user set scripted code line number and the ratio of total code line number, title and the size of motion picture files in the matching degree of keyword, website in Website page and the initial weight of these five factors of ratio of the size of animation file in the ratio of Website page file size, website and Website page size of tested website in the pagefile size of first known website, the page according to demand, it is ensured that the initial weight sum of five factors is 1；

Step 1-2, according to the numerical value of these five factors in website and every initial weight thereof, calculate first known website and initialize score；

Step 1-3, according to the numerical value of five factors in second known website and initial weight, calculate this website and initialize score；

Step 1-4, by calculating each factor scores and the ratio of total score in first known website, try to achieve five new weights of five factors respectively；

Step 1-5, recalculate the score of second known website according to five new weights, and the initial score of the above-mentioned score regained Yu second website is done difference；

Step 1-6, judge the difference obtained in step 1-5 whether approximate 0, the most then complete the determination that weights are final, and according to first known website or five factors of second known website, and then the threshold value of five each scores of factor in acquisition website, iterative process terminates；Otherwise, the weights obtained are substituted into first known website, repeats step 1-4 and step 1-5；

Step 2, network address according to website to be measured gathers the pagefile size of this website, scripted code line number in the page, total code line number, the title of tested website, keyword in Website page, animation file size in motion picture files size and website in website, and according to the pagefile size of above-mentioned seven tested websites of parameter determination, scripted code line number and the ratio of total code line number in the page, the title of tested website and the matching degree of keyword in Website page, the size of motion picture files and the ratio of Website page file size in website, the size of animation file and five factors of ratio of Website page size in website；

Step 3, will calculate obtain five factor scores respectively compared with respective threshold value, it is achieved the adjustment to website and webpage；

If the pagefile size score of tested website is more than threshold value, then reduce pagefile size；Otherwise show result；

If scripted code line number and the ratio score of total code line number are more than threshold value in the page of tested website, then reduce scripted code amount in the page；Otherwise show result；

If the title of tested website and the matching degree score of keyword in Website page are more than threshold value, then optimize website keyword；Otherwise show result；

If the size of the motion picture files in tested website is more than threshold value with the ratio score of Website page file size, then reduce motion picture files quantity；Otherwise show result；

If the size of the animation file in tested website is more than threshold value with the ratio score of Website page size, then reduce animation file quantity；Otherwise show result.

The numerical value according to these five factors in website described in step 1-2 and every initial weight thereof, calculate website and initialize score, and formula is as follows:

$a=\underset{i=1}{\overset{5}{\mathrm{\Σ}}}(p_{1i}\×w_i)---\left(1\right)$

Wherein, a represents the scoring detected according to initial weight；p_1iRepresent the result of calculation of the i-th influence factor of first known website；w_iRepresent the weights corresponding to this influence factor.

The numerical value according to five factors in second known website described in step 1-3 and initial weight, calculate this website and initialize score, and formula is as follows:

$b=\underset{i=1}{\overset{5}{\mathrm{\Σ}}}(p_{2i}\×w_i)---\left(2\right)$

Wherein, b represents second known website detection score, p_2iRepresent the result of calculation of the i-th influence factor of second known website, w_iRepresent the weights corresponding to this influence factor.

Described in step 1-4 by seeking the ratio of each factor scores and total score in first known website, try to achieve five new weights of five factors respectively, formula is as follows:

$w_i^{\′}=\frac{p_{1i}\×w_i}{\underset{i=1}{\overset{5}{\mathrm{\Σ}}}(p_{1i}\×w_i)}---\left(3\right)$

Wherein, w '_iRepresent the i-th influence factor p after adjusting_1iCorresponding weights, w_iRepresent the weights corresponding to this influence factor.

The invention have the advantages that

A kind of method improving the searched rate in website of the present invention, by collecting factor and the pattern affecting search engine, linear iterative algorithm is utilized to be iterated obtaining weights, thus website ranking is drawn a mark that can quantify, and for the value of each influence factor with the product of weights, the reasonable proposal to web information flow can be proposed, reduce cost, improve the efficiency of website promotion.

Accompanying drawing explanation

Fig. 1 is the method flow diagram improving the searched rate in website of an embodiment of the present invention；

Fig. 2 is first each influence factor of website and the initial weight product coordinate diagram of an embodiment of the present invention；

Fig. 3 be an embodiment of the present invention second each influence factor of website and an iteration after weights product coordinate diagram；

Fig. 4 is the scoring threshold value coordinate diagram of an embodiment of the present invention.

Detailed description of the invention

Below in conjunction with the accompanying drawings an embodiment of the present invention is described further.

The development environment hardware configuration of the embodiment of the present invention is Windows operating system: Microsoft Windows 7；CPU:Intel Centrino2 internal memory: 2GB；Hard disk: 320GB.

A kind of method improving the searched rate in website, method flow diagram is as it is shown in figure 1, comprise the following steps:

Step 1, determine the factor affecting site search engine masses, including scripted code line number in the pagefile size of tested website, the page and the ratio of total code line number, the ratio of title and the size of motion picture files in the matching degree of keyword, website in Website page with the size of animation file in the ratio of Website page file size, website with Website page size of tested website, and by the way of iteration, determine the threshold value of above-mentioned five each scores of factor respectively；

The method using iterative algorithm in the embodiment of the present invention, data mapping (as shown in Figure 2 to 4) are obtained by the website of same key word ranking in detection search engine, abscissa is each influence factor, and vertical coordinate is the evaluation that this influence factor is corresponding.Then, by obtaining the difference of the maxima and minima of each influence factor as the initial value of weights shared by this influence factor.Result in data set being inputted in BP iterative algorithm and is trained, the weights of the BP iterative algorithm obtained after completing are added with initial weight, take the meansigma methods of the two, are finally balanced the process of weights, to avoid the apparent error caused due to incidental error.The key step of balance weight process is as follows:

Step 1-1, determine first known website, the present embodiment is chosen the website of Northeastern University, pagefile size p₁Scripted code line number and the ratio p of total code line number in=25/50=0.48, the page₂=12/250=0.048, title and matching degree p of keyword in Website page of tested website₃=0.8, the size of motion picture files and the ratio p of Website page file size in website₄=0, the size of animation file and the ratio p of Website page size in website₅The initial weight w of=0 five factor_i(i=1,2 ..., 5)=0.2 (as shown in table 1), meet the initial weight w of five factors_i(i=1,2 ..., 5) sum is 1；

Table 1

Wherein, initializing weights is system definition, does not have any practical significance, only the weights initial value in iterative algorithm.

Step 1-2, according to the numerical value of these five factors in website and every initial weight thereof, calculate website and initialize score；

Formula is as follows:

$a=\underset{i=1}{\overset{5}{\mathrm{\Σ}}}(p_{1i}\×w_i)=0.2656---\left(1\right)$

Wherein, a represents the scoring detected according to initial weight；p_1iRepresent the result of calculation of the i-th influence factor of first known website；w_iRepresent the weights corresponding to this influence factor.

The score of website ranking can be calculated by formula (1).This formula is linear iterative algorithm model, by using above iterative formula, by the simple change of initial value, can produce value is very abundant, difference is big weights and the vector machine of threshold values.Therefore initial value w is selected_i, then by continuous interative computation, it is possible to the weight of each influence factor of Query refinement, present invention utilizes the mode of pattern recognition, the unknown weight method Query refinement by iteration.

Step 1-3, according to second known website, the embodiment of the present invention uses software institute of Northeastern University website, the numerical value p of its five factors_j(j=1,2 ..., 5), p₁=1.02, p₂=0.1875, p₃=0.7, p₄=0.3, p₅=25.6；And the initial weight w in step 1-1_i(i=1,2 ..., 5)=0.2, calculate this website and initialize score；Second known website described in first known website, step 1-3 described in step 1-1 is that recall ratio is high, the website that retrieval ranking is forward.

Formula is as follows:

$b=\underset{i=1}{\overset{5}{\mathrm{\Σ}}}(p_{2i}\×w_i)=5.5615---\left(2\right)$

Wherein, b represents second known website detection score, p_2iRepresent the result of calculation of the i-th influence factor of second known website, w_iRepresent the weights corresponding to this influence factor.

Step 1-4, by seeking the ratio of each factor scores and total score in first known website, try to achieve five new weights of five factors respectively；If being 0 after the weighed value adjusting of some influence factor, then with totally deduct be not 0 weights after mean allocation；

Formula is as follows:

$w_i^{\′}=\frac{p_{1i}\×w_i}{\underset{i=1}{\overset{5}{\mathrm{\Σ}}}(p_{1i}\×w_i)}---\left(3\right)$

Being computed, result is as follows:

w₁'=0.361

w₂'=0.036

w₃'=0.602

w₄'=0.0005

w₅'=0.0005

Wherein, w '_iRepresent the i-th influence factor p after adjusting_iCorresponding weights.

Step 1-5, recalculate the score of second known website according to five new weights, and the initial score of the above-mentioned score regained Yu second website is done difference, then remove absolute value；

Formula is as follows:

$s=\left|\underset{i=1}{\overset{5}{\mathrm{\Σ}}}(p_{2i}\×{w_i}^{\′}-b)\right|=4.75218---\left(4\right)$

Wherein, the appraisal result of second tested website after behalf revises weights and the difference of unmodified front appraisal result.

Step 1-6, judge the difference obtained in step 1-5 whether approximate 0, the most then complete the determination that weights are final, and according to first known website or five factors of second known website, and then the threshold value of five each scores of factor in acquisition website, iterative process terminates；Otherwise, the weights obtained are substituted into first known website, repeats step 1-4 and step 1-5；

According to each influence factor's score value condition calculating targeted website, it is judged that optimize direction present in website, it is provided that reasonably advise.

S_i=p_2i×w″_i (5)

Wherein, S_iIt is the score threshold of the i-th factor of second known website, w "_iFinal weights for i-th factor；

w″₁=0.104

w″₂=0.39

w″₃=0.286

w″₄=0.109

w″₅=0.201

Choosing second website is standard web site, is computed

S₁=0.10608

S₂=0.73125

S₃=0.2022

S₄=0.0327

S₅=5.1456

Step 2, network address (the homepage http://epub.cnki.net/kns/default.htm of National IP Network in using in the embodiment of the present invention) according to website to be measured gathers the pagefile size of this website, scripted code line number in the page, total code line number, the title of tested website, keyword in Website page, animation file size in motion picture files size and website in website, and according to the pagefile size of above-mentioned seven tested websites of parameter determination, scripted code line number and the ratio of total code line number in the page, the title of tested website and the matching degree of keyword in Website page, the size of motion picture files and the ratio of Website page file size in website, the size of animation file and five factors of ratio of Website page size in website；

C_i=(p '_i·w″_i) (6)

Wherein, C_iRepresent the score of targeted website i-th influence factor；w″_iRepresent the weights after determining；p′_iI-th factor for detected website；

C₁=0.1248

C₂=0

C₃=0

C₄=0

C₅=0

Step 3, will calculate obtain five factor scores respectively compared with respective threshold value, it is achieved the adjustment to website and webpage；

If the pagefile size score of tested website is more than threshold value, then reduce pagefile size to 50K；Otherwise show result；

The page is excessive, and page-downloading speed can be caused slow.Meanwhile, part searches engine only captures the partial content of the page, thus cannot obtain intended ranking effect.

If scripted code line number and the ratio score of total code line number are more than threshold value in the page of tested website, then reduce scripted code amount in the page；Otherwise show result；

Too much script can disturb the reptile of search engine to be analyzed web page contents, virtually reduces Keyword Density, affects the distribution of webpage weight.

If the title of tested website and the matching degree score of keyword in Website page are more than threshold value, then optimize website keyword；Otherwise show result；

If the title of website and page keyword are not inconsistent, it will reduce and change the website degree of association in this keyword search.

If the size of the motion picture files in tested website is more than threshold value with the ratio score of Website page file size, then reduce motion picture files quantity；Otherwise show result；

Search engine can not resolve the content in dynamic picture well, causes the disappearance that web site contents is obtained by search engine.

If the size of the animation file in tested website is more than threshold value with the ratio score of Website page size, then reduce animation file quantity；Otherwise show result.

Owing to search engine is the most unfriendly to Flash, it is impossible to finding out the link wherein hidden, this definitely to be stopped.Although Google now begins to search the content in record Flash, but important link such for leading boat, it is definitely can not to make of Flash, nor intuitively, speed of download is also slow, no matter for search engine user, is all the most disagreeableness.

Thering is provided solution to every influence factor's score, feed back to user, user is by the website of above-mentioned suggestion amendment oneself.

Claims (4)

1. the method improving the searched rate in website, it is characterised in that: comprise the following steps:

Step 1, determine the factor affecting site search engine masses, including foot in the pagefile size of tested website, the page This lines of code and the ratio of total code line number, tested website title with in the matching degree of keyword, website in Website page The size of motion picture files and the size of animation file in the ratio of Website page file size, website and Website page size Ratio, and by the way of iteration, determine the threshold value of above-mentioned five each scores of factor respectively；

Step 1-1, user set according to demand in the pagefile size of first known website, the page scripted code line number with The ratio of total code line number, title and the motion picture files in the matching degree of keyword, website in Website page of tested website Size and the ratio of Website page file size, website in the size of animation file and the ratio these five of Website page size because of The initial weight of element, it is ensured that the initial weight sum of five factors is 1；

Step 1-2, according to the numerical value of these five factors in website and every initial weight thereof, calculate first known website and initialize Score；

Step 1-3, according to the numerical value of five factors in second known website and initial weight, calculate this website and initialize score；

Step 1-4, by calculating each factor scores and the ratio of total score in first known website, try to achieve five factors respectively Five new weights；

Step 1-5, recalculate the score of second known website according to five new weights, and by above-mentioned regain The initial score with second website is divided to do difference；

Step 1-6, judge the difference obtained in step 1-5 whether approximate 0, the most then complete the determination that weights are final, And according to first known website or five factors of second known website, and then obtain five each scores of factor in website Threshold value, iterative process terminates；Otherwise, the weights obtained are substituted into first known website, repeats step 1-4 and step 1- 5；

Step 2, network address according to website to be measured gather scripted code line number, total generation in the pagefile size of this website, the page Animation file in motion picture files size and website in keyword, website in code line number, the title of tested website, Website page Size, and according to scripted code line number in the pagefile size of above-mentioned seven tested websites of parameter determination, the page and total code row The ratio of number, tested website title and Website page in the matching degree of keyword, website the size of motion picture files with The size of animation file and five factors of ratio of Website page size in the ratio of Website page file size, website；

Step 3, will calculate obtain five factor scores respectively compared with respective threshold value, it is achieved the tune to website and webpage Whole；

If the pagefile size score of tested website is more than threshold value, then reduce pagefile size；Otherwise show result；

If scripted code line number and the ratio score of total code line number are more than threshold value in the page of tested website, then reduce foot in the page This size of code；Otherwise show result；

If the title of tested website and the matching degree score of keyword in Website page are more than threshold value, then optimize website keyword； Otherwise show result；

If the size of the motion picture files in tested website is more than threshold value with the ratio score of Website page file size, then reduce Motion picture files quantity；Otherwise show result；

If the size of the animation file in tested website is more than threshold value with the ratio score of Website page size, then reduce animation file Quantity；Otherwise show result.

The method of the searched rate in raising website the most according to claim 1, it is characterised in that: described in step 1-2 according to this net The numerical value of five factors of standing and every initial weight thereof, calculate website and initialize score, and formula is as follows:

$a=\underset{i=1}{\overset{5}{\Σ}}(p_{1i}\×w_i)---\left(1\right)$

Wherein, a represents the scoring detected according to initial weight；p_1iRepresent the i-th influence factor of first known website Result of calculation；w_iRepresent the weights corresponding to this influence factor.

The method of the searched rate in raising website the most according to claim 1, it is characterised in that: described in step 1-3 according to second The numerical value of five factors in known website and initial weight, calculate this website and initialize score, and formula is as follows:

$b=\underset{i=1}{\overset{5}{\Σ}}(p_{2i}\×w_i)---\left(2\right)$

Wherein, b represents second known website detection score, p_2iRepresent the meter of the i-th influence factor of second known website Calculate result, w_iRepresent the weights corresponding to this influence factor.

The method of the searched rate in raising website the most according to claim 1, it is characterised in that: described in step 1-4 by asking first In individual known website, each factor scores and the ratio of total score, try to achieve five new weights of five factors respectively, and formula is as follows:

$w_i^{\′}=\frac{p_{1i}\×w_i}{\underset{i=1}{\overset{5}{\Σ}}(p_{1i}\×w_i)}---\left(3\right)$

Wherein, w '_iRepresent the i-th influence factor p after adjusting_1iCorresponding weights, w_iRepresent the power corresponding to this influence factor Value.