Research On Semantic Integration And Application Of Distributed Data Based On Linked Data

The development of the society cannot be separated from the information, theInternet today is the main route of information transmission. With the socialdevelopment, the information on the Internet is being increased at an increasinglyrapid rate. Faced with the vast amounts of information resources, their organizationand usage is particularly important. With the popularity of the network, and theincreasing amount of information, the current information organization has showngreat drawbacks. Therefore, the next generation of Internet (Semantic Network) hasbeen proposed and is evolving. It is intended to organize the information from thesemantic level, in order to increase the usage degree of information.Linked data is a new technology based on semantic web. This paper mainlystudies the application of linked data in the semantic integration of distributed data.The main research content is reflected in the "semantic" and "data integration". Andthe―semantic‖means ontology in semantic web; the "data integration" means the datapublishing based on the ontology.First, this paper studies the semi-automatic construction of domain ontology.Traditional ontology construction is mainly by hand. It will consume a lot ofmanpower and resources. Therefore, the automatic and semi-automatic ontologyconstruction gradually becomes an important research topic. This started later indomestic. And the flexibility of the Chinese grammar and vocabulary increases thedifficulty of Chinese ontology automatic construction. This paper analyzes the maindifference between Chinese and English, as well as the current major ontologyconstruction method. At the same time, we make a detailed study on Chinese ontologysemiautomatic building, and made a detailed description of the key technical issues,such as the corpus acquisition, corpus preprocessing, the term acquisition, ontologyconcept and relation extraction. In this paper we also made related experiments whichproved that the present method has practical usability.Then, this paper studies the publication of the linked data. Link data is based onontology, and aims to build a network of data between different data sources, so it canrealize data integration. This paper also summarized different ways of releasing linkeddata, according to different demands. Finally, we published the e-commerce data aslinked data as an example, based on the ontology constructed previously.