Bottom-up Approach of Modeling Human Decision Making for Building Intelligent Agents

摘要

Background/Objectives: Cognitive modeling of decision making with reference to the cognitive architectures and base theories is an ideal approach for building intelligent agents. This work is to conceptualize the modeling process as a bottom-up approach to build cognitive agents. Methods: Among the existing cognitive architectures, four system level architectures which are in similar nature are being sampled and different factors affecting the decision making scenarios reflected in the architectures are closely reviewed. Base theories of human cognition are also adopted for the chosen architectures to strengthen the modeling process. Findings: LIDA and CLARION are the two cognitive architectures found similar in symbolic and connectionist nature and are open architectures for modeling cognitive processes like high level decision making. The cognitive base theories are found suitable for modeling decision making with these architectures. On this way modeling process is to be done in a bottom-up fashion to build intelligent agents. Application/Improvements: The two analogous approaches with LIDA and CLARION will provide number of cognitive models of decision making. On implementation of these different models, diversified agents can be generated and their performance will be studied empirically.