# Cognitive architecture and behavioral model based on social evidence and resource constraints

## TL;DR

This paper introduces a cognitive architecture that models human behavior using social evidence and resource constraints, aiming to guide AI development aligned with human values.

## Contribution

The novel contribution is a hybrid knowledge-based cognitive architecture integrating symbolic and sub-symbolic knowledge with resource-aware inference and memory systems.

## Key findings

- The architecture can model human behavior and guide AI agents to align with human values.
- It supports practical applications in decision support systems and psychological modeling.
- The model can be applied to individual and group behavior analysis for various domains.

## Abstract

The cognitive architecture presented in this paper is expected to be able to explain certain aspects of human behavior, guide the development of artificial intelligence agents, and align the behavioral patterns of the latter with the former. The architecture is based on the principle of social proof or social evidence, including the principle of resource constraints. It includes the concept of a hybrid knowledge graph that encompasses both symbolic and sub-symbolic knowledge. This knowledge is divided into functional segments for fundamental, social, evidential, and imaginary knowledge, and is processed by an inference engine and a memory storage system that are aware of and manage resource constraints. The architecture and behavioral model derived on its basis are expected to be used to design artificial intelligence agents and decision support systems that are consistent with human values and experiences based on the alignment of their belief systems, capable of implementing decision support systems for practical applications. It can also be proposed for modeling human behavior individually or in a group, for psychological treatment, online security, and community management.