CN103838797A - Method for optimizing mobile search engine - Google Patents

Abstract

The invention discloses a method for optimizing a mobile search engine. The method comprises the following steps of designing a mobile search engine framework, establishing a URL list, editing a translator, and designing a WAP interface. Due to the fact that a mobile module is added on an existing Internet search engine framework according to the current situation of the mobile search engine, a mode for establishing the mobile search engine by HTML resources is provided. According to the mode, HTML web pages captured by web spiders are processed in a centralized mode, theme information extraction is conducted on the HTML web pages, then the theme information is converted into WML web pages which can be identified by a mobile phone and are stored in a WML snapshoot library, when a user click item record to check a specific web page, a system can not directly link the web page on the Internet but link a WML web page snapshot corresponding to the web page, and the requirement for mobile search of the user is met. In actual application, the mode is used for successfully establishing a mobile search engine facing the life service field and covering the catering, entertainment and yellow page information of nearly 40 cities in China.

Description

A kind of mobile search engine optimization method

Technical field

The present invention relates to development of Mobile Internet technology, particularly a kind of optimization method of mobile search engine.

Background technology

Search engine refers to according to certain strategy, uses specific computer program to gather information from internet, after information being organized and is processed, and for user provides retrieval service, the system by information display relevant user search to user.Now, along with the innovation of wireless communication technique and popularizing of mobile phone, mobile Internet access becomes development trend gradually, inquires about the demand of clothing, food, lodging and transportion--basic necessities of life information in order to meet user whenever and wherever possible, how to set up mobile search engine, becomes the focus of mobile network's application.Mobile Internet access is subject to the restriction of mobile phone terminal and transmission bandwidth, and pure html text only has minority intelligence type to support, most of mobile phones are only identified the language of wap protocol mark, as WML or xHTML.But the network information is mainly expressed with html language, the resource-constrained of WAP, cannot provide enough information to crawl the WAP page as the mobile search engine of information source merely.Therefore, how to break through restriction, make mobile phone client also can search the magnanimity information that derives from HTML, become one of subject matter of mobile search.For using mobile phone to browse html page, general method is to add a WAP gateway, in the time that mobile phone sends the request of browsing html web page, first reads this webpage by gateway, and converts it into corresponding WML, re-sends to mobile phone.This mode is also the current spread path that universal search engine is expanded to mobile search engine.But the mode of this real time translation, obviously performance and the bandwidth requirement to gateway is higher.

The present invention is directed to the present situation of mobile search engine, on the framework of existing internet search engine, add mobile module, propose a kind of HTML of utilization resource and set up the mode of mobile search engine, the html web page which captures by focusing on Web Spider, the html web page that Web Spider is captured is translated processing, translated into the snapshots of web pages of WML form, generated the snapshots of web pages of WML language, met user's mobile search demand.The mobile search engine of setting up with this technology, does not need the support of real time translation gateway, can expand easily existing search engine system.In actual applications, make successfully to set up in this way a mobile search engine towards service for life field, covered food and drink, amusement and the yellow page information in nearly 40 cities, the whole nation.

Summary of the invention

According to the present situation of mobile search engine, a kind of mobile search engine optimization method is proposed, comprise the following steps:

A, design mobile search engine framework

Comprise the following steps: this search engine framework is also made up of searcher, index, searcher, four parts of user interface, also has mobile module, as mobile search engine,

It comprises three parts:

Translater, the HTML page that spider is captured is converted into WML page;

WML snapshots of web pages storehouse, preserves the WML page after transforming;

WAP interface, by the user interface of mobile phone access;

B, set up url list

Deposit the webpage grabbing in web page library, and all hyperlink on webpage are deposited in url list;

C, editor's translater

Translater has home page filter, subject information filters and three parts of translation;

C.1 home page filter

First catalog page is filtered, will not translate, count and the ratio that links number according to the text section of webpage, carry out the character of paging, deposit index database in;

C.2 subject information filters

Extract the Topic relative part of webpage, select the tree-model of the STU-DOM that does not rely on information source,

Using the table of webpage, tr, div and tbody label node as piecemeal node, for the local correlation degree Local Correlativity for choice of a piece) and context dependent degree Contextual Correlativity weigh; Local correlation degree determines by piece internal chaining and content, and its computing formula can be expressed as:

$\mathrm{LinkCount}\left({\mathrm{STU}}_i\right)=\underset{j=1}{\overset{N}{\mathrm{\Σ}}}\mathrm{LinkCount}\left({\mathrm{STUC}}_{\mathrm{ij}}\right)$

$\mathrm{CountentLenth}\left({\mathrm{STU}}_i\right)=\underset{j=1}{\overset{N}{\mathrm{\Σ}}}\mathrm{ContentLength}\left({\mathrm{STUC}}_{\mathrm{ij}}\right)$

$\mathrm{LocalCorrelativity}\left({\mathrm{STU}}_i\right)$

$=\frac{\mathrm{LinkCount}\left({\mathrm{STU}}_i\right)}{\mathrm{CountentLenth}\left({\mathrm{STU}}_i\right)}$

Wherein, ContentLength and LinkCount represent respectively word number and the link number in piece, j sub-block of expression;

Context dependent degree determines by piece internal chaining and father's piece content, and its computing formula can be expressed as:

$\mathrm{Contextual\; Correlativ\; ity}\left({\mathrm{STU}}_i\right)$

$=\frac{\mathrm{LinkCount}\left({\mathrm{STU}}_i\right)}{\mathrm{CountentLenth}\left({\mathrm{STU}}_{\mathrm{Pi}}\right)}$

Wherein, STU _pirepresent STU _ifather node;

The design's regulation local correlation degree threshold value is 2, and the threshold value of context dependent degree is 70;

C3. HTML is transformed to WML:

In the time that HTML piece transforms, first to remove the element that WML cannot process, as labels such as style, front, script; Then, set up the mapping table that html tag and WML label transform, according to relation list, HTML be converted into the readable WML of mobile phone,

The text cannot a screen display on mobile phone showing, need to carry out paging processing, and deposit in the snapshot storehouse of WML;

D, design WAP interface

WAP interface is the man-machine interaction query interface taking mobile phone as carrier; Adopt WML or xHTML language design; Content on design WAP is as far as possible terse: in the list page of Search Results, entry number is no more than at most ten.

Compared with prior art, the present invention has following beneficial effect:

1, the present invention can break through restriction, makes mobile phone client also can search the magnanimity information that derives from HTML, for mobile search provides information widely.

2, the mobile search engine that the present invention sets up with this technology, does not need the support of real time translation gateway, has departed from performance and the higher problem of bandwidth requirement to gateway, can expand easily existing search engine system.

Brief description of the drawings

The present invention has accompanying drawing 2 width, wherein:

Fig. 1 is mobile search engine system frame diagram.

Fig. 2 is mobile search interface schematic diagram.

Embodiment

A, design mobile search engine framework

Comprise the following steps: the same with general search automotive engine system, this search engine framework is also made up of searcher, index, searcher, four parts of user interface, add mobile module, make it to become and extend the mobile search engine of expanding out, it comprises three parts:

Translater, the HTML page that spider is captured is converted into WML page;

WML snapshots of web pages storehouse, preserves the WML page after transforming;

WAP interface, by the user interface of mobile phone access.

Basic framework as shown in Figure 1.

B。, set up url list

First the present invention is started by Web Spider, regularly automatically starts and captures internet site, deposits the webpage grabbing in web page library, and all hyperlink on webpage are deposited in url list.

C, editor's translater

Due to mobile search engine need fast, directly, refining Query Information is returned to user, but in the webpage that spider captures, not only exist part without theme page; And, even there is the page of theme conventionally also to have the irrelevant information of a large amount of and theme.Therefore directly translate and be not suitable for, according to the feature of mobile search, according to the feature of mobile search, translater is designed to home page filter, subject information filters and three parts of translation.

C.1 home page filter

First catalog page is filtered, will not translate, count and the ratio that links number according to the text section of webpage, carry out the character of paging, deposit index database in.Index carries out word segmentation processing by oneself through the web document capturing, and the position occurring in webpage by word and frequency computation part weights, then deposits word segmentation result in index database.

C.2 subject information filters

Extract the Topic relative part of webpage, selection does not rely on the tree-model of the STU-DOM of information source, using label nodes such as the <table> of webpage, <tr>, <div> and <tbody> as piecemeal node, weigh for local correlation degree for choice (Local Correlativity) and the context dependent degree (Contextual Correlativity) of a piece.Local correlation degree determines by piece internal chaining and content, and its computing formula can be expressed as:

$\mathrm{LinkCount}\left({\mathrm{STU}}_i\right)=\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}\mathrm{LinkCount}\left({\mathrm{STUC}}_{\mathrm{ij}}\right)$

$\mathrm{CountentLenth}\left({\mathrm{STU}}_i\right)=\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}\mathrm{ContentLength}\left({\mathrm{STUC}}_{\mathrm{ij}}\right)$

$\mathrm{LocalCorrelativity}\left({\mathrm{STU}}_i\right)$

$=\frac{\mathrm{LinkCount}\left({\mathrm{STU}}_i\right)}{\mathrm{CountentLenth}\left({\mathrm{STU}}_i\right)}$

Wherein, ContentLength and LinkCount represent respectively word number and the link number in piece, j sub-block of expression.

Context dependent degree determines by piece internal chaining and father's piece content, and its computing formula can be expressed as:

$\mathrm{Contextual\; Correlativ\; ity}\left({\mathrm{STU}}_i\right)$

$=\frac{\mathrm{LinkCount}\left({\mathrm{STU}}_i\right)}{\mathrm{CountentLenth}\left({\mathrm{STU}}_{\mathrm{Pi}}\right)}$

Wherein, STU _pirepresent STU _ifather node.

The design's regulation local correlation degree threshold value is 2, and the threshold value of context dependent degree is 70.Html web page is carried out to subject information extraction.

C.3 HTML is transformed to WML

In the time that HTML piece transforms, first to remove the element that WML cannot process, as labels such as <style>, <front>, <script>.Then, set up the mapping table that html tag and WML label transform, according to relation list, HTML is converted into the readable WML of mobile phone, subject information is changed into the WML page that mobile phone can be identified, larger for word length, the text cannot a screen display on mobile phone showing, also needs to carry out paging processing, and deposits in the snapshot storehouse of WML.

D, design WAP interface

WAP interface is the man-machine interaction query interface taking mobile phone as carrier.Adopt WML or xHTML language design.Content on design WAP is as far as possible terse: in the list page of Search Results, entry number is no more than at most ten.In the time that user passes through WAP interface Query Information, first searcher carries out word segmentation processing to the information of user's input, and retrieve all records that comprise term, by calculating webpage weight and correlativity, query note is sorted, carry out set operation, the summary info that finally extracts each webpage feeds back to inquiring user.But in the time that user clicks bar record and watches concrete webpage, different from internet search engine, system can directly not link this webpage on internet, but links the corresponding WML snapshots of web pages of this webpage.

The method for designing according to the present invention, has developed service for life field mobile search engine www.zhaocha.mobi.It is to improve on the basis of original internet search engine www.zhaocha.com.cn, realizes effect as shown in Figure 2.

Claims (1)

1. a mobile search engine optimization method, is characterized in that: comprise the following steps:

A, design mobile search engine framework

Comprise the following steps: this search engine framework is also made up of searcher, index, searcher, four parts of user interface, also has mobile module, as mobile search engine,

It comprises three parts:

Translater, the HTML page that spider is captured is converted into WML page;

WML snapshots of web pages storehouse, preserves the WML page after transforming;

WAP interface, by the user interface of mobile phone access;

B, set up url list

Deposit the webpage grabbing in web page library, and all hyperlink on webpage are deposited in url list;

C, editor's translater

Translater has home page filter, subject information filters and three parts of translation;

C.1 home page filter

First catalog page is filtered, will not translate, count and the ratio that links number according to the text section of webpage, carry out the character of paging, deposit index database in;

C.2 subject information filters

Extract the Topic relative part of webpage, select the tree-model of the STU-DOM that does not rely on information source,

Using the table of webpage, tr, div and tbody label node as piecemeal node, for the local correlation degree Local Correlativity for choice of a piece) and context dependent degree Contextual Correlativity weigh; Local correlation degree determines by piece internal chaining and content, and its computing formula can be expressed as:

$\mathrm{LinkCount}\left({\mathrm{STU}}_i\right)=\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}\mathrm{LinkCount}\left({\mathrm{STUC}}_{\mathrm{ij}}\right)$

$\mathrm{CountentLenth}\left({\mathrm{STU}}_i\right)=\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}\mathrm{ContentLength}\left({\mathrm{STUC}}_{\mathrm{ij}}\right)$

$\mathrm{LocalCorrelativity}\left({\mathrm{STU}}_i\right)$

$=\frac{\mathrm{LinkCount}\left({\mathrm{STU}}_i\right)}{\mathrm{CountentLenth}\left({\mathrm{STU}}_i\right)}$

Wherein, ContentLength and LinkCount represent respectively word number and the link number in piece, j sub-block of expression;

Context dependent degree determines by piece internal chaining and father's piece content, and its computing formula can be expressed as:

$\mathrm{Contextual\; Correlativ\; ity}\left({\mathrm{STU}}_i\right)$

$=\frac{\mathrm{LinkCount}\left({\mathrm{STU}}_i\right)}{\mathrm{CountentLenth}\left({\mathrm{STU}}_{\mathrm{Pi}}\right)}$

Wherein, STU _pirepresent STU _ifather node;

The design's regulation local correlation degree threshold value is 2, and the threshold value of context dependent degree is 70;

C3. HTML is transformed to WML:

In the time that HTML piece transforms, first to remove the element that WML cannot process, as labels such as style, front, script; Then, set up the mapping table that html tag and WML label transform, according to relation list, HTML be converted into the readable WML of mobile phone,

The text cannot a screen display on mobile phone showing, need to carry out paging processing, and deposit in the snapshot storehouse of WML;

D, design WAP interface

WAP interface is the man-machine interaction query interface taking mobile phone as carrier; Adopt WML or xHTML language design; Content on design WAP is as far as possible terse: in the list page of Search Results, entry number is no more than at most ten.

Images