A Principle-Based Approach for Natural Language Processing

In natural language processing, an important task is to recognize various linguistic expressions. Many such expressions can be represented as rules or templates. These templates are matched by computer to identify those linguistic objects in text. However, in real world, there always seem to be many exceptions or variations not covered by rules or templates. A typical approach to cope with this situation is either to produce more templates or to relax the constraints of the templates (e.g., by inserting options or wild cards). But the former could create many similar case-by-case templates with no end in sight; and the latter could lead to lots of false positives, namely, matched but undesired linguistic expressions. Thus, the flexibility of rule matching has troubled the natural language processing (NLP) as well as the artificial intelligence (AI) community for years so as to make people believe that rule-based approach is not suitable for NLP or AI in general. On the other hand, fine-grained linguistic knowledge cannot be easily captured by current machine learning models, which resulted in mediocre recognition accuracy. Therefore, how to make the best out of rule-based and statistical approaches has been a very challenging task in natural language processing.