Research On The Fusion Of Massive Heterogeneous Minority Cultural Resources

In the information age with huge data,how to find the information needed by users from scattered,hidden and heterogeneous data,how to elevate information from data level to knowledge level,and how to complete high-quality knowledge service with the generation of new knowledge are particularly important.Yunnan’s ethnic minority cultural resources are very rich,but these resources are scattered and heterogeneous in different degrees,which hinders the effective dissemination and utilization of ethnic cultural resources,and the heterogeneous knowledge fusion can be from scattered and heterogeneous knowledge.The source has unearthed knowledge that has not yet been discovered,implied and useful,which provides a good method and means for knowledge acquisition,knowledge organization and knowledge utilization of ethnic cultural resources.Therefore,this paper conducts in-depth research on the integration of massive heterogeneous minority cultural resources,including the following three aspects:(1)Constructing a heterogeneous knowledge base of ethnic cultural resources.Firstly,the cultural resources of ethnic minorities are preprocessed.In the process of processing text resources,HMM-based word segmentation and part-of-speech tagging method and BiLSTM-CRF model-based named entity recognition method are proposed.Then,syntactic analysis technology is used to extract the relationship between entities.Finally,the extracted knowledge is stored in the form of RDF/XML and constitutes a heterogeneous knowledge base of ethnic cultural resources.(2)Propose a knowledge fusion algorithm for ethnic cultural resources.Aiming at the slow speed of the fusion algorithm of mass cultural resources in the traditional mode,this paper proposes a knowledge fusion algorithm based on attributes and rules in Hadoop platform and MapReduce parallel computing framework,and validates its effectiveness.In the similarity calculation,the algorithm proposed a similarity calculation method based on the improved concept attribute,and identified contradictory knowledge and removed redundant knowledge of extracted triples of ethnic minority cultural resources through ontology alignment,so as to obtain better fusion results.(3)Design and implement a visualization system for minority culture visualization based on knowledge fusion.On the basis of system demand analysis and functional design,a platform for knowledge fusion of ethnic cultural resources was established.The platform mainly realized the Internet content crawling of minority cultural resources,the HMM participle of minority text resources,the part of the minority text resources,and a few the national text resource triad extraction and the integration of minority resources knowledge have five functions.