CAJ | 학술논문

针对目前没有通用藏文搜索引擎的情况，提出一种基于领域本体的面向主题的藏文信息爬取策略。策略首先根据藏文和藏文编码的特点，判断出藏文Web并进行映射转换。然后通过构建领域本体，进行概念相似性计算，得到领域主题相关性网页。实验结果表明，该方法能够自动发现领域内的主题相关资源，并且比基于关键词的藏文网页信息搜索能得到更好的查全率和查准率，可以作为藏文Web语义搜索研究的参考。
침대목전몰유통용장문수색인경적정황，제출일충기우영역본체적면향주제적장문신식파취책략。책략수선근거장문화장문편마적특점，판단출장문Web병진행영사전환。연후통과구건영역본체，진행개념상사성계산，득도영역주제상관성망혈。실험결과표명，해방법능구자동발현영역내적주제상관자원，병차비기우관건사적장문망혈신식수색능득도경호적사전솔화사준솔，가이작위장문Web어의수색연구적삼고。
Aiming at the situation that there is no general Tibetan research engine at present,the thesis puts forward a domain ontology-based topic-oriented Tibetan information search strategy.The strategy first judges the Tibetan Web and carries out mapping transform according to the features of Tibetan and Tibetan coding.Then through constructing domain ontology it calculates the concept similarity,and obtains the correlation webpages of domain topics.Experimental results indicate that the method can automatically find the topic-related resources in the domain,and can get better recall rate and precision rate than the keywords-based Tibetan webpages information search. Therefore,it can be used as the reference for the study of Tibetan Web semantics search.