| 摘要: |
| 仅仅在一层知识上进行推理的推理模型有许多不足之处，例如，推理模型本身不能解释其行为的合理性（从观察者角度看），推理模型与外界进行信息交互时往往不能模拟人类的信息交互过程等．本文介绍一种深表层推理模型（即ＨＥＬＩ），它在一定程度上克服了上述不足，同时，本文还将讨论ＨＥＬＩ的变种及实例化等问题． |
| 关键词: 推理模型，解释，信息获取，知识获取，深层推理，表层推理 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| HELI: A DEEP-SHALLOW INFERENTIAL MODEL |
|
Cao Cungen
|
| Abstract: |
| Inference on a single level may suffers a number of deficiencies, for instance, the rationality of the inference can not be accounted for inside the level, and information acquisition is not human-oriented. This paper introduces a deep-shallow inferential model, called HELl. This model removes some bugs in single-level inferential models.Also, variations of HELI are presented. |
| Key words: Inferential model, explanation, information acquisition, knowledge acquisition, deep inference, shallow inference. |
