Using Ontology-Driven Methods to Develop Frameworks for Tackling NLP Problems

In this paper, we present the meta-tooling framework named TAISim that can be used both as a developer‟s tool for creating NLP systems and as a NLP learning environment, which allows helping students to construct NLP systems by example in a flexible way. TAISim enables the end user to combine different components of a typical NLP system in order to tackle specific NLP problems. We use ontology-engineering methods to accumulate meta-knowledge about the system construction and about users‟ activities to control the process of development and using the NLP system. Thanks to ontologydriven methods TAISim can be modified and enriched with additional information resources and program modules by means of a high-level interface. Additionally, we demonstrate how the using of meta-ontology helps us to improve TAISim to tackle ontology design automation problems.