ANALOG: A Knowledge Representation System for Natural Language Processing

The general task of work in knowledge representation and reasoning (KRR) is, trivially, the representation of knowledge. The domain for which the representation task has been undertaken is, often, secondary to the description of the overall task as an exercise in knowledge representation. Thus, the task of representing and using mathematical, common sense, visual, language-based, and logical knowledge is \lumped" together under the rubric of KRR (for examples of the diversity of goals of KRR see [12]). I will argue that in natural language processing and understanding there is a is a strong argument for the language structure and use mediating any attempt at representation. The task of knowledge representation for natural language processing and understanding is a knowledge-intensive one. To understand a natural language sentence a NLP system must, minimally, be able to represent the content of the sentence in the language of representation. These representation languages have been largely unmotivated with respect to the natural language they may be representing. In [12] only six of the twenty-twoKRR systems presented are driven by natural language processing concerns.