Ontology Storage,Query Expansion And Mapping Evaluation In Big Data Environment

With the development of information technology,we are about to step into the 5G era and it also makes the production of product information more rapid and the generated amount of information even greater everyday.The trend of the Internet of Things,allows products in different fields to interact with information and be cross-border integration,which in turn makes the product increases more information of itself and involves the information of related products.The type of information becomes more diverse and more complicated.These series of changes have led to the leap in the number of product ontology in different areas,as well as the variety of features,and have also brought significant challenges to ontology research in related fields.Firstly,the expansion of the number and the variety of product ontology has made the traditional file storage and relational database storage for ontology is not easy,and the large-scale ontologies have also brought problems such as long response time and complicated queries on the query.Secondly,users are no longer satisfied with basic requirements such as storage and query,and have higher expectations for the display of query results.But now,product ontology information overload makes it difficult for different users to find their favorite ontology information.Users are bored with the results of the same query,and the demand for personalized queries is increasing.Finally,the development of the Internet of Things makes the mapping of ontology between different domains more frequent and urgent,and users are more concerned about the accuracy of mapping relationships between ontologies.But the current mapping system evaluation criteria are relatively simple,it is difficult to consider and compare the ontology mapping system in all directions.Therefore,in the context of big data era,thinking from the user's point of view,the research on ontology mainly has four major problems,such as "limited storage capacity","complex and slow querying","not pertinent to query results",and "single standard of mapping system evaluation".These four questions are based on the user's point of view,containing almost the whole ontology lifecycle except the ontology construction,and they gradually advance at the level,so this article starts related research based on these four problems.Then we hope to achieve"scale" of ontology storage,"fastness of ontology query,"personalization" of query expansion,and"systematization" of mapping evaluation.Aiming at the pain point of large-scale ontology storage,this paper proposes the idea of storing RDF ontology and OWL ontology in HBase database respectively.For RDF ontology,this paper explores trials of traditional three storage modes based on relational database in HBase.It also proposes an improved storage strategies,and do a summary and comparison between them.For OWL ontology,this article also proposes a set of HBase-based storage logic,and exploring its feasibility from the perspective of the query.Combining with the characteristics of HBase itself,it optimizes situation of the "limited storage capacity" and "complex and slow queries",and realizes the "scale" of ontology storage and the "fastness" of ontology query.Secondly,aiming at the problem that "the query results are not well-targeted",this paper expands the query and shifts the perspective from the vocabulary,structure,semantics and other aspects of the ontology itself to user habits,proposes a complete web log processing logic process,and applies these logics to realize the transition from universal ontology query expansion to personalized ontology query expansion,which enables the user to perform more targeted ontology query in the big data environment.Finally,this paper proposes a set of four perspectives on the basis of synthesizing current academic research results in this area,including the vocabulary level,the structure level,the semantic level,and the usage level.It also has concretely described and quantified the indicators of each level in hopes of improving the situation of "single mapping system evaluation criteria" on the ontology and completing the "systematization" of mapping evaluation.