The Empirical Study Of Knowledge Discovery Based On Ontology

With the development of the storage and application technology of the database, enterprise data generated during operation is also to increase in geometric progression. This would require massive amounts of data to get to meet the specific needs of data to provide decision support for enterprise managers. Data Mining and Knowledge Discovery (KDD) is generated in this context. The world KDD was first performed in August 1989 in the United States held in Detroit, in the first International KDD Conference. Fayyad defined "KDD" as that is om the data sets to identify valid, novel, potentially useful and ultimately understandable patterns of non-trivial process. A complete architecture of knowledge discovery is composited by data sources, data storage, data mining engine and front-end tool, in which the main contents of the data storage layer is the data warehouse. The establishment of a data warehouse is the process of heterogeneous data integration, but also the foundation and core of knowledge discovery.The "father" of datawarehouse Inmon proposed a definition widely accepted in his book "Building the Data Warehouse" in 1991:Data warehouse is a subject-oriented, integrated, and relatively stable, reflecting the historical changes of the data set used to support management decisions. It can be seen from the basic definition of a data warehouse thar subject-oriented is the most important elements of the data warehouse, but also the biggest difference. with the traditional database that application-oriented. Because of this feature of data warehouse, knowledge discovery has the realization of the semantic characteristics in human-computer interaction, which can be semantic knowledge modeled to allow the data source be found in the computer reasoning and judgments, to achieve complex data sources by data mining and knowledge discovery.Ontology was a concept originally derived from the philosophical. In the field of artificial intelligence, Grubber of The United States Stanford University Knowledge Systems Laboratory firstly proposed a widely accepted definition of ontology:"Ontology is a conceptual model of a clear specification. " As can be seen from the definition, ontology describes the conceptual model of the real world exists, and its essence is used to implement as well as between human-computer interaction between machines and machines. As a basis for semantic modeling, ontology introduced into the engineering knowledge discovery among heterogeneous data integration can help to resolve the problem, and the use of RDF (S) described in XML to build the underlying data sources contribute to knowledge discovery.This article is an empirical research, mainly based on ontology building data warehouse, and then in the data warehouse on Knowledge Discovery which completed the following series of tasks: Second-hand housing to the area of real-world terminology of object-oriented analysis;Using the OWL language to create a second-hand housing domain ontology; Second-hand housing by filling the field of ontology instances way, heterogeneous data gathered into a unified framework for ontology being formed XML formatted data source; Pairs of heterogeneous data sources in data integration and knowledge fusion operations; Analysis and design of the field of second-hand housing data warehouse; For ETL, an instance of the ontology being imported into the data warehouse; A simple multi-dimensional data of the data mining experiments.