Research And Implementation Of Scientific Data Analysis System Based On Information Fusion

Web2.0era means that there are vast amounts of complex information on the Internet. The information published by institutions and individuals is so trivial that it makes people confused when screening information. Similarly, the work processing scientific data also faces the challenge. The researchers hope to know the personal research level, the research status in a certain field or find the most likely partners. All these need the support of data. Therefore how to get effective scientific information from multiple sources is our concern.With the development of Semantic Web, ontology languages like RDF and OWL have appeared and opened a new world of information fusion. Ontology is a kind of information integration technology and it has widely used in knowledge system and knowledge representation. Ontology building and merging technologies have their unique advantages in information fusion. Thus more and more people have realized its importance.In this paper, the main work includes four aspects:(1) This paper proposes the method for scientific research information acquisition and data preprocessing and design information extraction algorithm for formatted text.(2) Based on ontology technology, this paper puts forward information fusion method, including ontology construction algorithm and ontology fusion algorithm. Considering the single environment cannot deal with large-scale data processing, this paper also proposes a parallel ontology construction and fusion approach adapted to MapReduce framework. Through the comparing test between single and parallel environment, we analyze the advantages and disadvantages of the two methods.(3) Based on Knowledge Graph, this paper makes the inference rules and calculates the importance of entities, therefore extend the knowledge coverage.(4) Based on these proposed key technologies and combined with the user requirements, this paper gives an overall framework of scientific data analysis system and implements the prototype system. The system has functions such as information extraction, data fusion, analysis and display for scientific research information. The practical application shows that the system has certain practical value.