Research On The Knowledge Element Ontology Method Of Knowledge Representation For Scientific Literature

In the open interconnected era of ubiquitous networks,especially in the era of big data,a large number of articles have grown dramatically and users are drowned in the knowledge ocean of scientific literature.After users obtain a large number of scientific articles from traditional retrieval systems,they still have to manually extract required knowledge points by whole document reading and knowledge screening.That procedure of knowledge acquisition consumes a lot of time and energy of researchers,which results in a low effectiveness of scientific research output.In addition,it requires a high level of professional knowledge of users,which forms a solid barrier of knowledge acquisition.The essential reason for above-mentioned dilemma of knowledge acquisition is the limitation of the current knowledge description and representation methods for scientific literature: firstly,the major representing content is numerous external features(such as author,press,etc.)and few content features(such as title,abstract,etc.)of scientific literature;secondly,ultimate representing objects are coarse-grained documents rather than its inner fine-grained knowledge points.In order to solve that dilemma and satisfy their growing requirement of a fast,accurate and intelligent procedure of knowledge acquisition,this dissertation,from the basic link of knowledge description and representation of information retrieval,studies the method of knowledge element ontology for knowledge representation.This dissertation,firstly,systematically analyzes the textual structure of scientific literature at home and abroad and obtains an uniform knowledge structure of scientific literature;secondly,provides a knowledge element ontology method of knowledge representation for scientific literature,transfers knowledge points with relatively complete semantic features of scientific literature into fine-grained knowledge units with a unified structure;thirdly,evaluate validity of the knowledge representation method based on knowledge element ontology of scientific literature by two aspects,structural evaluation by indicators and functional evaluation by the FaCT++ reasoner;finally,utilizes the PHP programming language and MYSQL database software to design an new search system embedded in that knowledge representation method and compares retrieval results of the new search engine with those of existing traditional literature search engine to reveal the superiority of finegrained knowledge retrieval process of the designed search system embedded in knowledge representation method and the validity of knowledge representing for knowledge units of literatures.The knowledge element ontology method of knowledge representation for scientific literature describes,represents and organizes the knowledge units in scientific literature.This method offers a basic theory and method for digging out potential knowledge units of scientific literature,improving the efficiency of literature knowledge retrieval and promoting the knowledge acquisition,sharing and exchange among users in different fields.The designed fine-grained knowledge search platform will provide a practical foundation for realizing the fine-grained knowledge retrieval system in future.