Search Engine Semantic relevance judgment

Intermediary transaction http://www.aliyun.com/zixun/aggregation/6858.html ">seo diagnose Taobao guest cloud host technology Hall

According to the search engine to judge the original fingerprint algorithm we know that the search engine can put the text inside the page information, extract certain information, can be keywords, words, sentences or paragraphs and their weight in the page to form an information fingerprint or a specific feature, through this feature to distinguish between the two documents, However, this is only a very programmatic judgment of the search engine, which inevitably leads to a lot of errors in search results, because the association of two documents is not only decided by the text information, but also depends on the semantic association behind the information. And this point search engine has been in the effort, now Google's whole sentence search and Baidu's intimate search is semantic search, is the information behind the semantic Association of Mining, which can make the user's search results more intelligent, to provide more consistent with user needs search results.

Let's take a look at the following 2 sentences to see if there is a connection

"How much iphone5?"

"Will Apple price drop?"

Obviously, if we look at it, these 2 sentences are definitely related, although there is no public word between the two sentences. Although the second sentence of the "Apple" may be refers to the apple, but because the first sentence has "iphone5", we will be very natural to "Apple" as a mobile phone products, and only relying on traditional fingerprint algorithm will not have this effect. People can judge the semantics of 2 sentences by simple context, so how do search engines conduct semantic analysis to determine their relevance? This is about the semantic theme.

What is the semantic theme?

The semantic theme is the hidden theme in the text, and the word "apple" in the example above includes both Apple and the theme of the fruit. When we compare the first sentence, the Apple theme matches the theme that the iphone represents, so we think they are related.

A theme is a concept, an aspect, which is expressed as a series of related words, and a subject contains several words with higher probability. These words have a strong correlation with the subject, which collectively defines the subject. For a passage, some words may come from this topic, some may from that topic, for example an article if involves "the Baidu" This topic, then "Chinese search", "Baidu", "the search engine" and so on words will appear at the high frequency, but if involves "the Google" This topic, then " English search, "Google", "search engine" and so on will appear very frequently. and search engine This word also exists in Baidu and Google these 2 themes.

Search engine is relative to the semantic relevance of the document, in contrast to the 2 documents will be a summary of the document information analysis, if the 2 documents corresponding to the topic, it is certainly more relevant. The semantic association judgment is to be contrasted, for example, you search the "Apple" search engine can not determine whether you want the fruit or cell phone, will only be based on a search demand for most people to give results, a single word contained in a number of topics search engine can not determine the subject you need, only according to your long-term search habits to determine, So you need to add a modification, search "Apple phone", "Apple and Samsung which is good", so that the search engine can accurately give you the topic you need.

Semantic relevance analysis plays a ' supporting Role ' in keyword ranking

Search engine included the page, found that some keywords appear together a lot of times, can form a common topic, so that these words are summed up in semantic-related words. When these semantically related words appear together to form a topic, the relevance of the core keywords of the page plays an enhanced role, and the rankings are promoted.

It's empty? Wood-wood seo For example, if you search for "election", search engines may get textual information that includes "voting", "campaign" and "ballot", but the word "election" is likely to appear with little or no presence in these sources of information, In other words, semantic search can interpret the words and similar words of the keywords, thus enlarging the accuracy and relevance of the search content information and the user search terms, and improving the ranking of the core keywords, which is much stronger than the correlation of single word. For example, if you want to write an article about the Apple phone, if you add the word apple in the article, the effect is not as good as the appropriate insertion of the iphone, mobile phone, jobs and other words.

At the same time from the SEO keyword point of view, can also expand the page title and content of relevance, the content of added and the Title keyword topic of high degree of relevance of the Word can make the title and content of the Web page to strengthen the relevance. So in addition to the appropriate insertion of the core keywords in the article, you can add words related to the topic or similar to make the results more relevant to the user search term.

To sum up, the Semantic Topic Association judgment is a tool that can excavate the hidden information behind the language, and we can also serve the optimization and promotion of the website according to the semantic relevance judgment of the search engine. Do not reject the search engine, try to approach it, know it, understand it, we can better use it, search engine rules are there, use not to see how deep you know. Wood-wood SEO public micro-letter: Mumuseo

Author: wood-Wood seo http://blog.sina.com.cn/s/blog_c206a2c30101gdyx.html