Technical Research And System Implementation Of Water Knowledge Graph Construction

With the development of water affair resources information,there are many problems in water affair data,such as multi-source heterogeneity and huge amount of data.There are a lot of structured monitoring data and unstructured text data in beijing water bureau,and the data storage locations are various.Besides,there are also a lot of data of water affair on the network.These data need to be integrated at the level of object classification and data type in order to share the data further and make it easy for the water affair related people to query the data.Based on the current distribution of water affair data and the demand of user for comprehensive data,it is necessary to develop a model that can integrate a large number of heterogeneous multi-source data and apply it.With knowledge graph,data integration and knowledge visualization and information recommendation can be performed.Therefore,in the face of these situations,this paper needs to find a method for constructing knowledge graph for water affair data,so as to establish a water affair knowledge graph construction system to realize the application of knowledge graph in the business of water affair.In this paper,the technology of knowledge graph construction and application is fully introduced into this system.The main work is as follows:First of all,it is necessary to have a deeper understanding of the background and current status of the knowledge graph,the existing data characteristics and the application status of the knowledge graph.This paper presents a top-down and bottom-up knowledge mapping approach for structured data and unstructured data,and at the same time,it draws on the Wordnet thesaurus,Dbpedia dictionary,water industry standard documents and water expert knowledge and experience.The structured data uses the D2 RQ tool to transform the semantic information of RDF(Resource Description Framework)text.The unstructured data uses Jieba for text segmentation,links to the lexical information of CN-Dbpedia dictionary,and uses Jena tool to transform RDF text.The transformed RDF file is entity-aligned with the previously constructed knowledge graph of the water top-level to realize the linkage between structured data and unstructured text data.Finally,the water information recommendation is completed based on the constructed knowledge graph of water affair.In the process of entity alignment,this paper tries to use edit distance algorithm,Jaccard algorithm and euclidean distance algorithm to classify structured data.At the same time,LDA,LSI and TF-IDF are used to classify the unstructured data.The core codes of the algorithm are given respectively.Through the comparison experiment and result analysis,it is found that the evaluation results of the edit distance Algorithm and the LDA algorithm are better than those of other algorithms,so the validity and accuracy of the construction of water affair knowledge graph are verified.Based on the constructed knowledge graph of water affair,the semantic distance algorithm is used for water affair information recommendation.The results not only realize the sharing of water information,but also validate the accuracy of the knowledge graph,and add some semantic rules to make the results more reliable.In this paper,Xitrum framework is used to build the system of knowledge graph of water affair and its application.The main functions of the system and the functions of each sub-module are described in detail.Finally,the main page is displayed.The construction and application system of water affair knowledge graph in this paper relies on the national project of "Major Science and Technology Program for Water Pollution Control and Treatment of China".It integrates the data of the water affair in an all-round way,and can play a major role in future decisions and other events,and has practical significance.