Abstract - Indexed Vs. Unindexed Searching: From Security Classifications to Forensics

     Elizabeth Thede discusses two methods of information retrieval - indexed and unindexed searching. Indexed searching involves the creation of an index that identifies words and their places within a document. This allows subsequent searches to proceed at a fast rate since the data is organized. Unindexed searching proceeds without the creation of an index. The article also includes a discussion of which technique to use for specific projects and goals. Examples include adding security classifications to data, searching data outside the enterprise, searching forensically retrieved data and filtering e-mails.

     Both indexed and unindexed searching have their place in the enterprise. Indexed text retrieval is typically more efficient for uses such as general information retrieval and security classification systems. But unindexed searching too has its place - in outgoing email filtering, searching of live data sources like RSS news feeds, and sometimes in forensics. This article will attempt to explain which search technique to use when, and why.

Overview: Indexed Text Retrieval
     Indexing the inevitable millions of documents that any sizeable organization generates on shared file servers is the fastest way to facilitate data retrieval. An index will typically store each unique word in a document collection and its location within each document. Indexing also works with non - document