The RECON abstract only partially describes my thesis; more importantly in it I developed the metric distance technique for spectral classification, and showed it to be better than cross-correlation.
Other accessible articles describing my thesis work are: