Cassandra Search Indexes
by, 02-23-2012 at 06:10 PM (455 Views)
We can perform full text search in some application by reading whole content of the document and searching for the required data in document at acceptable speed. This operation is performed every time a query is executed. This approach is not feasible and recommended in cases where a huge amount of data is used for searching. Also almost no one is using this approach to search documents or database.
With increasing amount of data on the internet, we need to search millions or sometime billions of documents and still perform second searches.
In indexed based approach, an index maintains each term with its document ID in which this term is found. Also it maintains and stores other information that is required for searching and creating ranking of documents. It is also similar to a backwards process in which a word is mapped with multiple documents. Indexing is a process in which words are extracted with other related information and stored in the index’s architecture. After creating index, you can improve your search speed significantly.