Latent Semantic Indexing is the practice of a search engine gathering the documents that match the search using both the search keywords and related keywords. Such a technique provides more documents while keeping most of those documents to the proper subject.
In the old style of indexing or search engine optimization, only those articles or documents that contain the keywords noted in the search would be able to be pulled up using such a search. Using these techniques, more documents are pulled relating to the keywords, which makes the system appear more intelligent while providing more information to the user.
An example of this would be to use the keyword Dog Grooming. With traditional search engines, only those articles that containing the words dog grooming would be brought forward for notice. Using the LSI not only would the articles containing the words dog and grooming appear, but after them, words such as pets and care would also be keywords in other articles or documents.
While this Latent Semantic Indexing is a great tool, it should be noted that it does not understand the meanings of words. For example, Dog Grooming can eventually get words like care, and care can be care for the environment or care for the elder. These are not documents that one would be interested in if they are looking for Dog Grooming. Although, the LSI system is very effective at determining the appropriate documents to bring forward.
The manner that you achieve this matching of words is quite complex. The primary idea of this is to take a listing of every word in ever article, and go from there. First one would get rid of all of the articles, all of the pronouns, all of the adjectives, and all of the fancier words such as therefore. This leaves a series of content words that the system is able to understand and act upon.
The article Understanding Latent Semantic Indexing was Submitted by Romain Levesque through Articles.GetACoder.com network. Here's the additional information: Hi, I'm Romain Levesque Author of the  rticle |
|