论文部分内容阅读
门诊专家信息可以为需要就诊的居民提供一定的参考,通常这类信息分布于各个医院的门户网站,而且大多数需要通过查询才能获取。为了自动抽取医院门户网站的门诊专家信息,首先要解决查询接口判别的问题。针对此问题,根据查询接口的属性特征,构建了一种树状结构的领域模型,用于对查询接口进行分类判别,并进行领域关键词的填写。其次,对于返回的网页,需要进行噪声过滤,本文针对此问题,提出了分块重要度模型。实验结果表明,基于领域模型的查询接口判别方法比基于规则方法的准确率提高了10.83%,分块重要度模型的F1值比XPath方法提高了10.5%。
Outpatient expert information can provide some reference to residents who need medical treatment. Usually, this kind of information is distributed in various hospital portals, and most of them need to be queried. In order to automatically extract outpatient expert information from the hospital portal, we must first solve the problem of discriminating the query interface. To solve this problem, according to the attribute of the query interface, a domain model of tree structure is constructed, which is used to classify the query interface and fill in the field keywords. Secondly, for the returned web pages, noise filtering is required. In this paper, we propose a block importance model for this problem. Experimental results show that the query interface discriminant method based on the domain model improves the accuracy rate by 10.83% compared with the rule-based method, and the F1 value of the partition importance degree model increases by 10.5% compared with the XPath method.