Neural-Fuzzy Connectionist Semantic Memory Model

Abstract

This dissertation introduces a novel model for the semantic memory named ‘Neural-fuzzy Connectionist Semantic Memory Model’. Fuzzy Inference System (FIS) is used to implement this memory structure according to neural network with fuzzy logic concept. The proposed model is also supported with a Graphic User Interface (GUI), by which the user can simply form the question, and get the appropriate answer(s). The obtained results were quiet accurate and encouraging. The dissertation is subdivided into four chapters organized as follows: Chapter 1: It includes an introduction about memory categories, and semantic memory in general. The chapter also introduces the objective and the outline of this research. Chapter 2: In this chapter the semantic computing field is illustrated. Also, the semantic memory models are discussed from the psycholinguistic view. These models have different architecture according to the computer science. Chapter 3: This chapter introduces detailed explanations of how to formulate the research problem, and the methodology for the design of the proposed semantic memory model. Also, this chapter includes the designed system layers in detail.