A Knowledge Base Construction Towards Urban Public Infrastructure With Its Application In Condition And Tendency Monitoring

Nowadays,the construction of smart city is being applied to an increasing number of scenes.It turns out to be an extremely essential demand that the administration activities are supposed to be equipped with the intelligent ability.Public critical infrastructures such as roads,bridges and tunnels,transportation hubs,urban public areas,communities and the nursing home are the main scenarios of the smart city.The researches of the topic focus on analyzing the real-time data and making appropriate decisions based on the phenomena reflected by the problems raised from the operating condition which is collected from the sensors installed on the infrastructures as well.In the context of the smart city,there are a considerable sum of urban roads,including the certain forms composed of bridges and tunnels,which are the prime example of the public infrastructures.Therefore,we are required to further understand the regulation of the facilities.Entities and relationships will be constructed from the demanded knowledge contents screened out by mapping the existing expert base.Thus,the models which are designed to monitor situation on the basis of our knowledge graph will get improvements.This research enables to understand the perational rules of urban road facilities better as well as to provide powerful support to the construction of smart cities.This thesis will focus on the following points:(1)Our research,analyzing the feature about urban roads,turns to be a dataoriented project.In order to select the satisfied articles which are collected from expert knowledge bases and the standard literature,word vector method,combined with the semantic rules,is adopted in this paper which enables to calculate the similarity of various topics so that the data support can be provided to the knowledge graph.(2)The algorithm of word sense disambiguation based on cluster analysis is supposed to be applied in the process of data fusion from various sources so as to achieve knowledge entity alignment expressed in different forms.Experiments have been conducted to confirm the availability of the method which enables to improve the accuracy of the knowledge graph.(3)The state monitoring of the urban public traffic roads is selected for the application in this paper.The procedure is proposed that the risk model can be optimized by the guidance provided by the knowledge base.Moreover,the risk prediction and the cognitive inversion of urban road condition comes to be a reality so that the knowledge graph can be applied into the situation monitoring.