Figment Software Limited

We were requested by a client to provide a basic but complete Information Retrieval system to index ASCII text files and provide a search facility based on indices built by the system periodically. Fundamentally, this design is the same for all "Search Engines" (for example, Google).

The indexing component of the system was implemented in Java and employed an Information Retrieval concept called TF-IDF, which analyzes document content based on the frequency of terms within a document (excluding stop words such as "the" and "and"). The more frequent a term, the more the document is believed to be about that term.

A text-based user interface was sufficient for the client, but a rich graphical user interface could be added with ease, as could any other interface mechanism (for example, web-based).

The solution was both concise and complete, fulfilling the needs of the client. The indexing process was accurate within the limitations of the mechanism employed (dictated by the client budget). The search mechanism, performed on the index, was fast, efficient and accurate.

For further information, Contact Us.