Consideration Of Web Text Sensitivity To Water Environment

Despite the relative abundance of water resources in Fujian Province,with the rapid socio-economic development of the province,regional water supply and demand balances are very different,and pollution problems such as excessive discharge of pollutants,garbage disposal,and black and odor-laden water bodies in cities are prominent.This article takes Fujian Province as the scope of research,and based on the basic data,monitoring data,and statistical data of water environment,it adds online text data and uses"Internet thinking" and data mining technology to develop online text mining and sensitivity of water environment.Evaluation analysis and research help to help environmental protection management and decision-makers formulate scientific and rational plans for water environmental governance.It is of great significance for the construction of water environment ecological civilization.The main contents include:constructing a water environment network text recognition model based on Bayesian network method to determine the water environment network text and its pollution types;constructing a water environment sensitivity assessment model based on fuzzy analytic hierarchy process to explore the water environmental sensitivity of various districts and counties in Fujian Province The main influence factors and spatial distribution characteristics of various sensitive impact factors.The specific research content and achievements are as follows:(1)Adopting web-based web crawler technology to collect and acquire online text data related to water environment in Fujian province from April to June 2017 on the Internet,and to process the text segmentation and segmentation of the web text.The results show that the water environment The spatial distribution of related texts shows that the southeast coast is dense,the northwestern inland is sparse,the east is mainly distributed in densely populated and economically developed areas,and the west is mainly distributed along the trunk rivers;the water quality classification method is used for the April-June 2017 in Fujian province.Water quality monitoring data are evaluated.From the compliance rate of water quality monitoring indicators in various regions,the water quality compliance rate in Nanping City is the highest,and the water quality compliance rate in Putian City is the worst.(2)Construct a text recognition model of water environment network based on Bayesian network,that is,use the existing knowledge and experience to analyze the causality between network node variables,determine the Bayesian network structure,and adopt the Expected Maximum Estimation(EM)algorithm.Parameter learning is performed on the training sample data to determine the conditional probability distribution of each network node variable.Compared with other identification model methods,the model has achieved high verification results in terms of overall accuracy,confirmation rate,and recall rate,and it has features such as dynamics,high efficiency,stability,and interpretability of recognition results.Certain advantages can satisfy the recognition of text information and its pollution types in complex water environment networks.(3)The sensitivity evaluation model based on FAHP can intuitively reflect the water environment sensitivity and spatial distribution characteristics of various districts and counties in Fujian Province.From the pollution sensitivity index,the sensitivity of the province as a whole is higher in the southern region than in the northern region,especially in the southwest and southeastern regions.The high-sensitivity regions are scattered and scattered mainly in the middle and lower reaches of the Ting River,the lower reaches of the Jin River,and Long River;From the protection sensitivity index,the sensitivity of the province as a whole is higher in the northwest-south-central region than in the northeast-southwest region,and the high-sensitivity regions are scattered,mainly distributed in the tributaries of the upper reaches of the Min River,including Jian River,Mulan River,and Qiaolu River;According to the text sensitivity index,the sensitivity of the province as a whole is higher in the eastern-central region than in the western-central region,and the highly sensitive areas are concentrated in the lower reaches of the Min River.According to the analysis of the sensitivity indices,the overall water environmental sensitivity of the province decreased from southeast to southwest-north-central-northeast,and the high-sensitivity areas were mainly concentrated in the economically developed areas along the east coast and the rivers entering the sea were more obvious.