IRS13 Scientific Report No. IRS-13 Information Storage and Retrieval Table of Contents table of contents Harvard University Gerard Salton Use, reproduction, or publication, in whole or in part, is permitted for any purpose of the United States Government. TABLL iF CONTENTS (continued) Page II. continued A) Single Number Measures . B) Varying Cut-off Performance Curves C) Comparison of Single Number and Curve [OCRerr]. The Construction of Average Precision Versus Recall Curves . . . e . A) Averaging Techniques . . B) Cut-off Techniques e . . C) Extrapolation Techniques for Request Generality Variations * D) Extrapolation Techniques for Evaluation of Cluster Searching * . 5. Measures for Varying Relevance Evaluation 6. Measures for Varying Generality Comparisons 7. Techniques for Dissimilar System Comparisons and Operational Testing . . . 8. The Comparison of Specific and General Requests the Viewpoints of the "higher precision11 and "high recall" user . . * 11-6 * 11-7 Measures 11-13 * . 11-18 * . 11-18 * * 11-21 and 11-31 * 11-140 11-143 * 11-146 11-51 11-53 9. The Presentation of Data as Individual Request Merit 11-63 References e . . e . . * . 11-67 III. KEEN, E. M. "Search Matching Functions" 1. Introduction . . . . . e e 111-1 2. Matching Procedures used in Manual, Mechanized and Automated Systems . . . . 111-1 v