标题：A Novel Holistic-based Entity Unifying Method for Heterogeneous Data (Open Access)
作者：Zheng, Jiecai ;Lu, Zhaoxia ;Zhao, Xinxiao ;Li, Xueqing
作者机构：[Zheng, Jiecai ;Lu, Zhaoxia ;Li, Xueqing ] School of Sport Communication and Information Technology, Shandong Sport University, Jinan, China;[Zhao, Xi 更多
会议名称：2019 4th Annual International Conference on Information System and Artificial Intelligence, ISAI 2019
会议日期：May 17, 2019 - May 18, 2019
来源：Journal of Physics: Conference Series
摘要： © 2019 IOP Publishing Ltd. All rights reserved.">With the fast growing of the informatic technologies and extensive accumulation of informative sources. The information is gradually presented with more characteristics such as messiness and heterogeneity. Traditional entity unifying methods tend to have good measurement results for structured data with uniform structure and relevant attributes. For unstructured multi-sourced data, the entity similarity cannot be accurately measured. To overcome the defects of the traditional methods, this paper proposes an entity unifying method for heterogeneous data by calculating the entity similarity based on "relation". Extensive experiments on different types of datasets by have been conducted to demonstrate the superiority of our method.
© 2019 IOP Publishing Ltd. All rights reserved.