Describes a microcomputer-based string index format generated from concept networks using multiple search terms. Link weighting is highlighted as a method to decide which term appears first so the entry reflects the classification implied by the search specification. The computer software used for this system is briefly described. (Author/LRW)

Provides an interim report on a project whose goal is to provide subject access to the knowledge of mankind at least as well as the skilled reference librarian. Classification and concept formation are discussed in terms of four modules: diagnostics of information-seeking behavior, network interface, subject access, and graphic interface. (EM)