An efficient execution method for rule-based machine translation

A rule based system is an effective way to implement a machine translation system because of its extensibility and maintainability. However, it is disadvantageous in processing efficiency. In a rule based machine translation system, the grammar consists of a lot of rewriting rules. While the translation is carried out by repeating pattern matching and transformation of graph structures, most rules fail in pattern matching. It is to be desired that pattern matching of the unfruitful rules should be avoided. This paper proposes a method to restrict the rule application by activating rules dynamically. The logical relationship among rules are pre-analyzed and a set of antecedent actions, which are prerequisite for the condition of the rule being satisfied, is determined for each rule. In execution time, a rule is activated only when one of the antecedent actions are carried out. The probability of a rule being activated is reduced to near the occurrence probability of its relevant linguistic phenomenon. As most rules relate to linguistic phenomena that rarely occur, the processing efficiency is drastically improved.