Information excessive capture is finite: Index talking about search props up dependency technology

Core clew: Dependency, it is the key in searching engine to optimize, basically decide by the characteristic that searchs engine. Above all Web webpage amount has been achieved on the dimensions of billion, next, the search brings those who prop up user oneself to retrieve professional competence normally very finite.

Dependency, it is the key in searching engine to optimize. But the principle of search engine work to dependency, believe major SEOER understands to be being lacked. But the way that needs our mainstream to search engine technology only, you can know to search the pulse of engine times.

The generation that sort technology spends related basically is decided by the characteristic that searchs engine. Above all, the Web webpage amount that contemporary search engine can visit has been achieved on the dimensions of billion, even if the user just is searched among them very few one part content, the search engine that is based on full text to search a technology also can be returned by tens of thousands page. Even if these result webpages are user place need, the user also is impossible all pair of webpage once-over, the result webpage that can be interested in the user most so is put at in front, certainly will can enhance the satisfaction that searchs engine user to spend. Next, the search brings those who prop up user oneself to retrieve professional competence normally very finite, behavior retrieves in most general keyword in, the user is key person only commonly a few terms. For example, nearly 300 Spink users that ever waited for search engine to Excite had done an experiment to investigate, discover what average per capita inputs to retrieve a word to be 3.34. Domestic share scholar also has similar conclusion, discover 6 to the Chinese that 90% right-and-left users input retrieves individual character to be 2 ~ , and 2 words word is in the majority, occupy about 58% , it is 4 words word next (take 18%) and 3 words word about (occupy 14%) about.

Too little retrieval the retrieval that the word cannot convey an user truly in fact demand, and the user also does not undertake complex logic constructs normally, only quite little user undertakes Boer logic retrieves, restricted retrieval mixes advanced retrieval to wait for a method, only the retrieval of 5.24% is included in type Boer logic calculates accord with. The research of partial scholar of home also makes clear as a result, the user of about 40% cannot apply field retrieval correctly or retrieve 2 times, 80% right-and-left users cannot apply advanced retrieval correctly function, still discover the user lacks motivation to learn complex retrieval even skill, most user sends a hope to be able to be their tectonic and effective retrieval automatically at searching engine type. Because lack the online in the past retrieval retrieve personnel that what middling often has, accordingly, the retrieval with actual user behavior and the retrieval with ideal user behavior exists in fact difference, those who retrieve a result is dissatisfactory also be not strange. As a result of this characteristic, search engine must try the user most before wanted webpage result puts webpage result as far as possible, this spends sort algorithm to be in search engine related the webpage namely why very main reason.

Those who show level is relevant spend sort technology to basically have the following kinds: It is the means that is based on technology of traditional Information Retrieval, what it basically uses keyword itself to inquire a requirement to documentation and user in the important rate in documentation is relevant degree make measure, be like the frequency that the keyword in using a webpage appears and position. Generally speaking, the inquiry keyword that in retrieving the webpage documentation that give, contains number is more, dependency is bigger, and the area graduation of this keyword jumps over tall; at the same time, it is important to if appear in field of such as caption to wait,inquire a keyword locally, compare those who appear in text is relevant degree should big. 2 it is technology of analysis exceeding cable length, the representative search engine that uses this technology has Google and Baidu to wait. Compare with former photograph, the important rate that it is approved with the webpage serves as those who retrieve a result is relevant spend sort basis. Look from design thought, it pays attention to tripartite to be opposite more of this webpage approbate, if have bigger chain,the webpage that counts into the webpage just is the important webpage that gets be admissived extensively, and the traditional method according to keyword position and frequency is the form that ego of a kind of webpage approves only, devoid objectivity. There still are a few other ways finally, if define sort by user freedom regular define way oneself. The day net FTP of Beijing University searchs engine to use means of this kind of sort, it can let an user choose the specific sort index such as such as time, size, stability and distance to undertake to result webpage relevant spend sort. If collect fees,rank mode again, it regards a kind of when search engine main gain as the method, wide in the large search engine that has network portal characteristic to use, but the objectivity that searchs a result at worrying about an influence, this kind of means is not their mainstream sort means, and show as a complement merely in paying cost to search a column.