Ontology-Based Model and Procedure Creation for Topic Analysis in Chinese Language

This paper focuses on setting up a methodology to create models and procedures for the functions and processes involved in topic analysis, especially for the business documentation in the Chinese language. Ontologies of different types are established and maintained containing the annotated and evolved knowledge. Extraction, transforming, and loading methods are adapted from approaches which are used to set up a data warehouse for standardized data models, exploited as the basis of a large variety of analysis. Topic discovery is conducted based on Latent Dirichlet Allocation for different usage. An interactive tool is implemented to support the proposed design and realistic demands.