Based On Multi-source Home School Communication Data Individual Feature Analysis Support Platform

At the beginning of the 21st century,new ways of communication between home and school appeared,such as mobile campus communication,Unicom home school and telecommunication wing school,etc.With the rapid development of the Internet,the home and school communication platform after another,the platform data between the system and semantic level of heterogeneity,there is no uniform standard,and other home communication platform between incompatible,is bound to cause new "data Island".At present,the integration of heterogeneous data sources based on ETL in traditional data warehouse is very cumbersome and difficult to deal with,which can not solve the problem of data sharing,reuse and semantic integration.The quality of data integration is poor.As the ontology has the ability to express conceptual semantics and reasoning accurately,ontology generation rules are used to guide ontology mapping,which avoids the shortcoming of traditional heterogeneous data source integration.Based on the above problems,this paper presents an ontology-based heterogeneous data source integration scheme in data warehouse.By constructing an application domain shared vocabulary database,marking the data source schema and applying the ontology generation process,an automatic tag data storage algorithm is proposed.Then,through the ontology mapping and reasoning process,the ontology semantics is detected and hidden hierarchical structure information is found.Automate ETL operations based on these build rules.And then apply the integration scheme to the actual scene of multi-source home school communication data.The case shows that the ontology-guided ETL process solves the shortcomings of the traditional process and improves the integration efficiency of the data source.By analyzing the plug-in architecture oriented to the domain requirements,this paper proposes a data analysis platform based on plug-in technology,which is designed and implemented according to the functional modularization principle.A multi-data source adaptation framework is introduced for multiple data sources.A dynamic data mining algorithm plug-in adaptation framework is proposed for the data analysis algorithm library,and the algorithm plug-in is dynamically loaded and managed.Finally,based on the data analysis support platform based on the plug-in technology,this paper puts forward the basic data and behavior of the main students,parents and teachers through the integration of the heterogeneous data sources based on ontology,Data analysis and mining of individual characteristics for the results of scientific management and decision-making,it is bound to the current pattern of education to bring immeasurable value.