IRS13 An Analysis of the Documentation Requests chapter E. M. Keen Harvard University Gerard Salton Use, reproduction, or publication, in whole or in part, is permitted for any purpose of the United States Government. x-l X. An Analysis of the Documentation Requests E. M. Keen 1. Introduction Conclusions from the experiments performed with SMART have so far been made on the basis of the average results achieved for a set of search requests. Studies presented in sections III, V, VI, VII, VIII of this report have presented such results, and given in addition some elementary data about the individual requests, and examples of detailed system performance The purpose of this study is to uncover still more of the variables and charac- teristics of the requests themselves in the context of the test environment and of the capabilities of the SMART system, in order to reveal basic prob- lems and suggest in[OCRerr]rovements to the system. The 35 requests associated with the 82 document collection in documentation are used because the author's knowledge of the subject field aids the investigation. 2. Request Preparation In section I, a brief description of the ADI documentation collec- tion, requests, and relevance decisions is given. The collection consists of 82 documents, both in abstract and short text form. Requests were prepared by two non-users of the system, and the relevance decisions were also made by these persons by examining every document for potential relevance. The request preparers were graduate students at Harvard University, one an engineer, and one an applied mathematician. Since neither person