Abstract. This paper elaborates how to use the formal concept analysis method to map contents in the database to the ontology, in order to provide the big data application with high-quality data source by virtue of integrating the database with Semantic Web. In recent times, a mass of data is stored in the relational database, but such data with low share usage fails to play its full role. On account that the big data application has grown by leaps and bounds, a large number of shared data is urgently needed. By mapping the data in the relational database into the ontology, the technology of Semantic Web can provide a lot of semantic data to the big data application, which is conducive to big data analysis and use. In this paper, the ontology is built by taking the formal concept as the intermediate model and converting the logic structure of database into Hasse graph and context table, and then combining with the domain knowledge. The ontology in the knowledge domain can be found from the database by applying the formal concept analysis method, which takes full advantage of logical structure information of the database and is beneficial for automation found by the ontology. Eventually, ontology method and problems found in the relational database by virtue of the formal concept analysis are summarized herein..

Abstract-The use of machine learning method is discussed herein to produce a corpus by domain knowledge ontology and conduct text classification according to the ontology of professional knowledge domain. Nowadays, a large number of literature materials have been accumulated in each professional field, and it is still in rapid growth. This constitutes a great challenge for researchers in various fields. To be specific, not only the workload in literature retrieval and reading is constantly increased, but also the work efficiency of the study is affected. In this paper, ontology is taken as the text feature extractor for storage, processing, classification and retrieval through ontology development tools Protégé, Jena and natural language processing tool NLTK, so as to facilitate the researcher for literature retrieval and reading. The advantage of this text classification method lies in that category structure is no longer a single tree structure, but instead, different categories may intersect and new category may be grouped by themselves.