Research Article
Time-Varying Water Quality Analysis with Semantical Mining Technology
@INPROCEEDINGS{10.1007/978-3-030-48513-9_29, author={Jun Feng and Qinghan Yu and Yirui Wu}, title={Time-Varying Water Quality Analysis with Semantical Mining Technology}, proceedings={Cloud Computing, Smart Grid and Innovative Frontiers in Telecommunications. 9th EAI International Conference, CloudComp 2019, and 4th EAI International Conference, SmartGIFT 2019, Beijing, China, December 4-5, 2019, and December 21-22, 2019}, proceedings_a={CLOUDCOMP}, year={2020}, month={6}, keywords={Pattern mining Sequence patterns Association rules Water quality forecasting}, doi={10.1007/978-3-030-48513-9_29} }
- Jun Feng
Qinghan Yu
Yirui Wu
Year: 2020
Time-Varying Water Quality Analysis with Semantical Mining Technology
CLOUDCOMP
Springer
DOI: 10.1007/978-3-030-48513-9_29
Abstract
Water resources is one of the most important natural resources. With the development of industry, water resource is harmed by various types of pollution. However, water pollution process is affected by many factors with high complexity and uncertainty. How to accurately predict water quality and generate scheduling plan in time is an urgent problem to be solved. In this paper, we propose a novel method with semantical mining technology to discover knowledge contained in historical water quality data, which can be further used to improve forecast accuracy and achieve early pollution warning, thus effectively avoiding unnecessary economic losses. Specifically, the proposed semantical mining method consists of two stages, namely frequent sequence extraction and association rule mining. During the first stage, we propose FOFM (Fast One-Off Mining) mining algorithm to extract frequently occurred sequences from quantity of water quality data, which can be further considered as input of the second stage. During the process of association rule mining, we propose PB-ITM (Prefix-projected Based-InterTransaction Mining) algorithm to find relationship between frequently occurred water pollution events, which can be regarded as knowledge to explain water pollution process. Through experimental comparisons, we can conclude the proposed method can result in flexible, accurate and diverse patterns of water quality events.