Document clustering, using macro retrieval techniques

An algorithmic approach for generating a loosely structured set of documents based upon references and citations is described and initially tested. The technique is based upon a small subset (or single relevant document) of known relevant docuemnts which form the basis of the network of structured references and citations. Suggestions for improvement and refinement point out potentially ripe research avenues for investigating the indentification of relevant docuemnts pertaining to emerging interdisciplinary subject areas.