作者: Graeme Hirst , Peter A. Heeman
关键词: Process management 、 Artificial intelligence 、 Affect (linguistics) 、 Inference 、 State (computer science) 、 Plan (drawing) 、 Action (philosophy) 、 Sanctions 、 Referring expression 、 Order (business) 、 Computer science
摘要: This paper presents a computational model of how conversational participants collaborate in order to make referring action successful. The is based on the view language as goal-directed behavior. We propose that content expression can be accounted for by planning paradigm. Not only does this approach allow processes building expressions and identifying their referents captured plan construction inference, it also allows us account clarify using meta-actions reason about manipulate derivation corresponds expression. To clarification goals arise inferred plans affect agent, we agents are certain state mind, includes an intention achieve goal currently considering. It mental sanctions adoption acceptance plans, so acts link between understanding generation.