A new type of adaptive-network-based computational verb inference system

Abstract

The adaptive-network-based computational verb inference system (ANCVIS) is a kind of computational verb inference system implemented in adaptive networks. In this paper, a new type of ANCVIS, which uses trend-based computational verb similarities, is presented, and its learning algorithm is derived. Simulations show that the proposed model can tune system parameters well and yield reasonable results based on computational verb rules and input-output data pairs.