Research on Heterogeneous Knowledge Fusion Algorithm under Big Data Environment

Yao ZHANG,Shu-yu LI,Yue TANG
DOI: https://doi.org/10.3969/j.issn.1673-629X.2017.09.003
2017-01-01
Abstract:In environment of big data,the integration of multi-source heterogeneous knowledge fusion has provided one of the most effec-tive means and methods for researchers to discover the implicit,valuable and undetected knowledge from a lot of knowledge sources that are dispersed and heterogeneous. Aimed at the shortcomings of the current knowledge fusion methods,based on investigations on them un-der the big data environment,the existing data fusion methods have been employed,which are transplanted to the knowledge fusion rea-sonably. A kind of algorithm for multi-source heterogeneous knowledge fusion is proposed. In order to further improve the quality of the acquiring knowledge,an improved algorithm based on the dynamic selection of knowledge source granularity is proposed to obtain the ap-propriate size of the collection of knowledge sources and the true and reliable knowledge as possible. Its experimental verification is con-ducted based on the experimental platform constructed by Hadoop and MapReduce framework. Experimental results show that it is effec-tive and feasible and effectively improves the performance of multi-source heterogeneous knowledge fusion algorithms.
What problem does this paper attempt to address?