Research On Query Technology Of Large-scale Biomedical Semantic Linked Datasets

In recent years, there are large amounts of Linked Data published on the Webwhich have semantics, and the biomedical datasets occupy a large proportion of all.As these datasets are distributed and query functions they provided are limited, whileeffective information are not fully mined and there is no comprehensive applicationsoffered to users. Therefore, the research on semantic queries of it combiningsemantics of RDF and biological significance has great theoretical significance andapplication value of engineering.Based on detailed analysis of11datasets including DBpedia, Sider, Diseasome,DailyMed, LinkedCT and so on, a platform of semantic queries with biologicalsignificance of multiple datasets is developed. In order to ensure the consistency ofRDF, algorithms using MapReduce for checking inconsistencies are designed, and wecheck the consistency of DBpedia which is the hub of the LOD and we giveexperimental results and solutions. We put forward the datasets relationship miningalgorithm and we draw out the relationship diagram of multiple datasets. According tothe relationship of these datasets, we design three semantic query problems: query ofdiseases, query of drugs according to the disease and query of side effects accordingto the drug. Taking Cassandra as the database, we distributed complete the dataloading using MapReduce. We put forward algorithms and implementation of thethree query functions using theory path query and instances queries presenteffectiveness and superiority of our platform.The platform for semantic queries based on large-scale biomedical datasets takesfull advantage of the technologies of Semantic Web and biomedical technologies. Notonly a practical and effective platform is offered to users but also we believe that thisway has certain guiding significance for developing the system of answeringquestions intelligently.