30 December 2006

Latent Semantic Analysis

I am still on vacation at the beautiful Texas coast, but I have found time to do some reading that I want to share. Remember, I am, first and foremost, an Internet researcher.

Latent Semantic Analysis (LSA) and Latent Semantic Indexing (LSI) are increasingly important concepts in search engine marketing (SEM) and search engine optimization (SEO).

LSA is a statistical technique for extracting and representing the similarity of meaning of words and passages by analysis of large bodies of text. In addition to recognizing keywords on a page, LSI searches all documents in the index, looking for similar terms. Documents that have many words in common are considered semantically close; documents with few words in common to be semantically distant. This method of classification is almost human in its scope. The LSI algorithm does not know the meaning of the words, but recognizes patterns that emerge from semantically close words.

Why should you, as a webmaster or marketer, care about LSA and LSI? Google changed its algorithm and has achieved success with displaying natural search engine results, versus keyword-driven search engine results. With keywords, the search engine results will include only those documents containing these exact keywords. With Latent Semantic Indexing, the search engine results will contain documents whose content is considered to be semantically close to your keywords.

To be continued . . .

Tags: , , , ,

No comments:

Related Posts with Thumbnails