A Dialogue Tutoring System Exploiting Extracting-Rule-Embedded-Template Structure and The Expanding Powers Of Buggy Rules

摘要

This paper develops a new tutorial dialogue system exploiting the extracting-rule-based extracting-ruleembedded-template structure and the expanding powers of buggy rules. The system is effective not only in simplifying, thus reducing otherwise time-consuming authoring task of template generation (|t,2|) but also in improving the system performance. The extracting-rule-embedded-template architecture introduced leads to system simplification and performance improvement firstly because it allows many different templates to be integrated into, or equivalently to be extracted from, a single extracting-rule-embedded-template by applying extracting rules assigned to some of its transitional nodes and secondly because the buggy rules introduced are capable of identifying and hence generating bugs from learners' erroneous responses automatically. The NLP (natural language processing) technology plays a key role in the system development because a parser is used to examine the syntactical structure of the learner's free format response while its semantic structure is examined by matching the respective response to the semantically equivalent paths of template databases prepared. The extracting rule-based and buggy rule-based extracting-rule-embedded-template structure is expected to play an important role in many applications including both written and spoken tutoring systems, voice-enabling call center or voice portal systems or in fact any system focusing more enhanced human computer interfaces, implementing more natural human computer interactions between the system and humans.