Exploring Probabilistic Top-k Dominating Query over Multiple Uncertain Data Streams

• Main Authors: Tien-Chun Wang, 王天駿
• Other Authors: Chuan-Ming Liu
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/w46b7n

碩士 === 國立臺北科技大學 === 資訊工程系研究所 === 105 === How to effectively derive the valuable data in Big Data is one of the important research issues in Data Science. The data we collect may be uncertain, imprecise, or a statistic due to the device or transmission error. In addition, the data may change as time evolves and we refer to such a data set as an uncertain data stream that has, velocity, veracity, and volume properties. This paper explores the top-k dominating query process on uncertain data streams and employs the parallel computation to facilitate the query process. The challenges include how to quickly update the result and reduce the computation cost for processing uncertainty. By referring the related existing papers for certain data, we provide an effective top-k dominating query process on uncertain data streams in terms of time and space and the provided approach can be parallelized easily. After discussing the properties of the proposed approach, we validate our methods through extensive simulated experiments. The experimental results indicates that our algorithms can avoid the unnecessary computation effectively and reduce lots of communication throughput between servers, thus achieving the objective of updating the results quickly.