The present report is the twenty-first in a series describing research in information storage and retrieval conducted by the Department of Computer Science at Cornell University. The report covering work carried out by the SMART project for approximately two years (summer 1970 to summer 1972) is separated into five parts: automatic content analysis (Sections I to V), dynamic document processing (Sections VI and VII), interactive feedback procedures (Sections VIII to X), automatic dictionary construction (Sections XI to XIV), and clustered file organization (Sections XV to XVII). This scientific report contains new experimental results dealing with a comparison between automatic and manual indexing methods (Sections I and II), proposals for dynamic collection control and automatic book retirement (Sections VI and VII), evaluation results for interactive feedback searches (Section VIII), experiments relating to the automatic construction and maintenance of dictionaries and thesauruses used for language normalization (Sections XI to XIV), and new automatic document clustering methods (Sections XVI and XVII). (Author)