A Spreading Activation Network for Automatic Generation of Thesaurus Relationships

Abstract

A spreading-activation-network for the generation of semantically-related terminology to obtain an output terminology such as synonyms is presented here. In contrast to the hitherto known networking models in computer linguistics, the network is automatically developed, namely from basic linguistic relationships. The principle of inhibition is applied in such a manner, that the interfacing edges having negative weightages serve towards indirect activation.