LSI term explained

Latent semantic indexing (LSI) is an algorithm search engines to analyze the web pages that takes into account not only key words/phrases on the page, but the presence of their synonyms, and words that are related thematically request. The algorithm is intended to in the top of issuance was the most relevant documents.

LSI is a special case of the latent semantic analysis (latent semantic analysis, LSA) in relation to the search engines. M can also be used in areas that require automated processing of a large number of documents:

• Classification;
• Indexing;
• Search.

The algorithm determines the semantic structure of the text and finds the implicit, latent connections between document elements. To do this, using a database of terms, semantically related words of the search query. Therefore, the system returns documents not only the exact occurrence of the query, but similar in meaning.

Thanks to the LSI algorithm, search engines have learned to recognize typographical errors, to distinguish between synonyms and homonyms, neologisms. 

When analyzing new text LSI is a robot compares it with the existing base of trust sites on the subject, takes into account the number of thematic words and synonyms, determines the semantic core of the page. Take into account the number of failures, close the page immediately after loading — the article should interest the reader. On this basis, the robot decides how to define the text in the results.

Thus, dropping out SEO content, stuffed with keys, but does not contain useful information, as well as texts from non-gambling f. As a result, the user receives a really useful article on the subject, not just SEO text for website promotion.