Research On Semi-automatic Domain Ontology Construction

Along with Internet's rapidly expansion and information technology's progress, network is not only the pure platform for communication, but also the information carrier of digital and informational resources. People search information that they need through search engines. However, information's independence and search engines' limitation of each professional field cause it unable to realize the information searching on the semantic level. For better realizing information sharing and exchange, we need the specialized-field ontology to instruct.Ontology is a very strange concept for many people, only philosophers understand this conception before. However, with the ontology is getting more and more widespread in the computer application domain, it has already become a hot spot in computer domain, more and more experts start to study ontology technology, we also know ontology very early as a graduate student in the school, and start to research on ontology. An outstanding ontology has a great help in many ways of computer domain, at first, ontology construction becomes the most important area of technology, how to use existing tools reasonable, how to improve existing algorithms, how to integrate each method mutually, how to raise efficiency and the accurate-rate of ontology construction, these questions are which to be solved urgently. This article is a research on ontology semi-automatic construction.Ontology constructs is very complex and tedious process, it needs many professional fields' experts to participate. Although, for the present, ontology-engineering tools is mature relatively, but the majority works are still completed manually that are tedious and laborious, which bring difficulties for ontology construction, also the hindrance for knowledge extraction.Primary coverage: This article is a research on methods of semi-automatic ontology construction, proposes a ontology semi-automatic construction flow based on text statistics. Basic steps are as follows: Firstly, construct a core ontology; Secondly, deal with texts' statistical information, extract texts' domain related key words in the ways of taxonomic approach and the extraction method, produce ontology conceptions candidate collections; Thirdly, judge domain terminology and core ontology concepts' semantic relations by Wordnet, add the extracted terminology to the core ontology automatically, realize the core ontology's expansion; Finally, produce the final ontology through artificial revision and consummation.Main research goal: Though the research on semi-automatic ontology construction, find a practical and feasible ontology semiautomatic construction method. This method can not only raise the efficiency of ontology constructs, reduce persons' work load in the construction process, also guarantee ontology's quality. Certainly, the final goal of the research on construction method is to develop a complete, highly effective and correct ontology constructed system, enables it to be widespread applied in ontology domain as other mature ontology tools.Experiment's harvest: Through the research on ontology, I have a more comprehensive understanding on ontology knowledge, including the concept, meanings and classifications of ontology, network ontology language OWL, ontology construction principles, constructing methods, tools and so on. The grasp of these knowledge has not only very great help for ontology construction, also provide a rationale foundation for the further application of ontology technology; This article proposed a ontology semi-automatic construction method based on text statistics through study and research on semi-automatic ontology construction method. Through the tests and confirmations to each step, confirm this method's feasibility and accuracy. This semi-automatic ontology construction method reduces person's work load in construction process, raised ontology construction's efficiency, and guaranted ontology's qualityExperiment's summary: For time and scientific research level's limits, the methods and functions developed are limited, but this was still a good attempt in the research and a practice of ontology construction, and built a good foundation for the continual study and research on ontology, also provided power and confidence of innovation of ontology construction methods and research and breakthrough of ontology construction tools.Prospect and forecast: Along with computer technology's development, ontology technology will manifest its more and more important value. As researchers in this direction, we should continue to be harder, not only absorb the overseas vanguard technology, study the overseas advanced method, practice and use; more importantly, also through unstoppable study and practice, accumulate experience, raise ability, develop independent ontology tool, propose more outstanding ontology construction method, promote our country's ontology research level to the advanced rank of the world.