An increasingly important AI frontier is the ability to represent worldviews, culture, values, and other nuanced structures, and to simulate the effects of these on perception, emotion/affect, judgment, and opinion formation. Such information, however, is notably difficult to model and represent, due to its fine-grained, diffuse nature. Reasoning is also highly challenging in these domains. This paper presents a novel 'EnergyBased' Knowledge Representation formalism (INTELNET) ideal for modeling, fusing, and reasoning about nuanced semantics, cultures, affects, and worldviews. It then introduces the integrated COGVIEW conscious/unconscious psychological simulation framework operating on top of INTELNET and advances a detailed example within the suicide terrorism domain.Applications include intelligent reasoning systems, humanitarian missions, cultural simulations, knowledge engineering, language processing, anti-discrimination and prejudice reduction, terrorism reduction, and norm change efforts, among others.
I. CULTURES, BELIEFS, CONCEPTS, AND MORE: INTELNET ENERGY-BASED KR (EBKR) AND THECOGVIEW FRAMEWORK The ability to create agents that can reliably integrate and simulate information regarding worldviews, including culture, values, and other conceptual structures represents an increasingly important frontier in artificial intelligence.[1], [2], [3] Worldviews mediate perception, but are notoriously difficult to model and to represent, in part due to their nuanced, diffuse natures. In computational systems, moreover, it is difficult to define, bound, and reason about whatever 'it' is that culture and worldviews consist of. Beyond this, effects of human cognition add another layer of complexity. This paper presents a new knowledge representation formalism, INTELNET, expressly targeted at nuanced data generated by systems grounded in and arising from the human cognitive capacity. INTELNET is ideally suited to worldview modeling and the nuanced semantics underpinning natural language, enabling simulation of the effects of worldviews on reasoning, persuasion, and understanding.In order to facilitate enhanced reasoning capabilities, INTELNET seeks to model nuanced semantics, representing information with a symbolic opacity intermediate between that of neural networks and typical symbolic systems. INTELNET concepts are represented as distributed, interconnected networks wherein each part of a network operates in concert with others to define a concept and to model the meaning of a particular semantic domain.Beyond nuance, computer worldview models are faced with certain 'quirks' of human psychological processing which lead complex conceptual systems to behave differently than they otherwise would if they were not being generated via human cognition. To this end, the paper presents a framework covering specific psychological phenomena that integrate with the INTELNET formalism in order to enable detailed simulations.INTELNET, together with the integrated psychological models discussed in this paper, are referred to a...