Data extraction and annotation based on domain-specific ontology evolution for deep web

Deep web respond to a user query result records encoded in HTML files. Data extraction and data annotation, which are important for many applications, extracts and annotates the record from the HTML pages. We proposed an domain-specific ontology based data extraction and annotation technique; we first construct mini-ontology for specific domain according to information of query interface and query result pages; then, use constructed mini-ontology for identifying data areas and mapping data annotations in data extraction; in order to adapt to new sample set, mini-ontology will evolve dynamically based on data extraction and data annotation. Experimental results demonstrate that this method has higher precision and recall in data extraction and data annotation.