The traditional methods of spectral text retrieval (FDS,CDS) create an index of spatial data and convert the data to its spectral form at query time. We present a new method of implementing and querying an index containing spectral data which will conserve the high precision performance of the spectral methods, reduce the time needed to resolve the query, and maintain an acceptable size for the index. This is done by taking advantage of the properties of the discrete cosine transform and by applying ideas from vector space document ranking methods.

Fourier Domain Scoring (FDS) has been shown to give a 60% improvement in precision over the existing vector space methods, but its index requires a large storage space. We propose a new Web text mining method using the discrete cosine transform (DCT) to extract use- ful information from text documents and to provide improved document ranking, without having to store excessive data. While the new method preserves the performance of the FDS method, it gives a 40% improve- ment in precision over the established text mining methods when using only 20% of the storage space required by FDS.