Research On Chinese-English-Mongolian Term Knowledge Graph Of Computer Field Based On WordNet

With the rapid development of information globalization,the interconnection of all things has become the development trend of the times.The information and data generated by this development trend are also rising in a straight line.These information and data can just be used as an effective data source for relationship analysis.If the traditional intelligence is to analyze each individual,then in today’s information age,the generation of interconnection of all things is not only to analyze each individual,but also to study the relationship between individuals has become a hot spot.For any project,as long as the relationship between individuals is analyzed,the shadow of the knowledge graph will appear.At the same time,with the rapid development of information globalization and network technology,the Internet has brought a new experience for people’s life,work and study,in the fields of information retrieval,machine translation,et al,the support of the knowledge base is needed,and the emergence of the knowledge graph has just solved this problem.In the process of information development,Mongolian information resources are also very rich,in practical application can be used directly,but relatively scarce,a large number of Mongolian information resources are simply transported to the Internet,which is extremely unfavorable for the effective transmission and utilization of Mongolian information resources.Therefore,it is urgent to construct the knowledge graph of Mongolian.The data in the underlying database of WordNet belongs to structured data,which roughly covers all the concept information.Therefore,this thesis selects the data in the underlying database of WordNet as the data source.The knowledge base based on WordNet is not rare,such as ontology database,semantic dictionary,Mongolian Noun Semantic Web and so on.In this thesis,WordNet is selected as the data source to improve the utilization of resources.Compared with unstructured data,it reduces the preprocessing of resources,saves time and speeds up the construction speed.This thesis studies the knowledge graph of Chinese-English-Mongolian terms in computer field based on WordNet:a.Firstly,the parameters that affect semantic similarity are analyzed.After the comprehensive analysis of the advantages and disadvantages of each parameter,the information content(IC)is defined as a parameter,and the improved IC value calculation model is used.When calculating semantic similarity,not only the information of concept nodes in WordNet,but also various semantic relationships need to be considered.After taking IC value as the parameter to calculate semantic similarity,four popular algorithms are analyzed systematically.Finally,this thesis selects the hybrid semantic similarity calculation method,the model verifies various semantic information,including distance,depth,IC,conceptual features and semantic relations.b.The construction of Chinese-English-Mongolian terms knowledge graph in computer domain based on WordNet is carried out according to the following process: Firstly,the computer domain terms are divided into several sub domains which are not included in each other by the method of manual intervention;Secondly,each sub domain is given a core concept,and the concept set of each core concept is extracted by using semantic similarity algorithm,Get the final concept set of the core concept;Thirdly,the final concept set is translated through Da-Er-Han Northeast Asia Language Translation Platform(which is supported by Niu Trans),and then the ChineseEnglish-Mongolian terms extracted from WordNet are integrated;Finally,the Neo4 j graphic database is used to visualize the constructed target domain knowledge graph.c.The construction of Chinese-English-Mongolian terminology knowledge graph in computer field based on WordNet provides experience for the construction of Mongolian knowledge graph,It also provides a reference for constructing a knowledge graph of minority languages.The purpose of showing the specific construction method and process is to provide some reference for the further improvement of Chinese-English-Mongolian trilingual knowledge graph.