well, the search engine for the search term is basically from the five aspects, and give you a summary here: to stop words, word segmentation, instruction processing, error correction, classification and treatment. Although this is not directly related to our website, but many.
, third instructions: usually we will input more words in the search, in fact, by default, the default search engine to a number of words according to the "and" word. For example, you search "lose weight" website ", the default search engine we input" weight loss website "to deal with, including" lose weight alone "or" content theory will not "return, but the actual situation often appear alone contains a word site. In addition, we need to deal with various advanced search instructions such as words, pluses and minuses etc..
A model of
fifth, classification treat: different words are generally treated differently, this is a feature of search engine. For example, we search keywords in general, return to the general web pages in the majority, and when we search a star’s name will appear a lot of news, video, pictures and other comprehensive search results. So, search engine to be judged according to the needs of users. This judgment is in the preprocessing stage.
search engine is actually WEB2.0, because of the need to handle user input data, so can very well and users interact. But for search engines, and not all words in the user input are directly used for ranking, but first after their treatment, the words can direct user ranking after the treatment. According to their summary and share of search engine is how to deal with the search words we enter. If not, we also welcome.
fourth, error correction: most of the time we would have the wrong input, Pinyin was homophonic more wrong, it is normal, but if the search engine without error correction, it will return a lot we don’t want to see the results. So, sometimes we even input spelling, search engine will be based on the right to prompt Chinese.
first, words segmentation is a major feature, whether we input long sentences or phrases, the search engine will put these words segmentation, will these search string cut into the word based keyword combination. So, in order to carry out the inquiry and comparison by himself after pretreatment has the database of all meet these words did not list returned to the user. The principle of segmentation and the segmentation of some similar pages.
second, to stop words: we are a lot of time in search of subjectivity, a lot of time to some of the words spoken words are used in the search words, but the words often to search content without substantial help, but we have formed the habit of input these words, so the search engine must to stop words, the maximum extent to ensure the accuracy of the search and correlation.