Functional Picture of Text Mining

How is Text Mining Diffrent then Web Search

Text mining is a process in which we are looking for data that we do not know exists such as patterns. We know there is data such as customer satisfaction comments but we do not know what usful information is contained within the data. Search on the other hand is a process where we know that the data exists but we are unsure of where and need to find it by pushing aside any data that does not relate to what we want.

How Text Mining Works.Edit

Text mining works by using algorithms to search through text in an effort to read the text like a human would read the text. Text mining software has to be able to understand languages more thoroughly then most other applications. It needs to under stand words, definitions and semantics.

It is worth noting that while the algorithms that text miners use are not perfect and are relatively new they are constantly being improved to return less errors, find more useful information and run quicker.

An example of why text mining can be more useful then just plain searching would be a looking through a news paper for medical articles. If our goal was to find medical articles regarding broken legs we would want the software to return articles about someone who “broke their leg” but not articles about someone who "just went broke" or about a thief who broke into the bank.