Font Size: a A A

Research On Construction Of Minority Festival Domain Ontology And Semantic Retrieval Model

Posted on:2018-07-31Degree:MasterType:Thesis
Country:ChinaCandidate:H KuiFull Text:PDF
GTID:2348330533465332Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
Minority culture is the spiritual wealth created and inherited in the process of national development.It has unique value,rich connotations and distinctive national characteristics.The traditional minority festival is an indispensable part of minority culture.However,due to the development of modern society and the strong impact of the mainstream culture,the development and utilization of the information resources of minority is insufficient.The traditional way of information dissemination makes the national characteristics,such as national customs,culture and religion,face the dilemma of inheritance.This thesis takes the Water-Splashing Festival as an example to achieve the semi-automatic construction of the ontology of minority festivals,and researches in the semantic retrieval model of the minority festivals.The advantage of the ontology technology is used to provide the technical basis for the inheritance,protection and propaganda of the minority culture.The main work of this thesis is as follows:1.This thesis summarizes the research status of semi-automatic construction at home and abroad,discusses the semi-automatic construction methods of domain ontology,puts forward a semi-automatic construction method of minority festival domain ontology,and constructs the initial core ontology,we construct a minority festival dictionary for texts segmentation through the experts guidance and consult a large number of documents.2.According to the needs of the field,this thesis obtains the texts through the web crawler technology,and exploits the SVM text classification technology to obtain the related texts of Water-Splashing Festival in the field of minority festivals.In the segment stage of texts preprocessing,we use the minority festival dictionary and the improvement of the feature selection and weight calculation is made,which improves the accuracy of classification.3.We extract the concepts and relationships from the classified texts,the method based on statistics is used to obtain the domain concept set in the concept extraction,based on the lexical features and grammatical analysis technology,the tree structural SVM is employed to extract the conceptual relationship.With the Jena API and Protégé,we complete expansion and refinement of the ontology to obtain the final domain ontology.4.In order to verify the practicality of the domain ontology,the detailed design and conception of the retrieval of the minority festivals domain ontology are put forward,and the initial semantic retrieval test model is constructed,which provides the theoretical basis and feasible research for the follow-up application of the ontology.
Keywords/Search Tags:Minority festival, Ontology, Semi-automatic construction, Text classification, Relational extraction
PDF Full Text Request
Related items