通过梳理搜索算法中出现的各种特征因子,提出一种更全面的特征因子分类方法,按其自身特点将特征因子划分为6类,分别是文字内容特征、Web拓扑特征、主题分布特征、文档类型特征、统计特征和时效性特征,并描述各特征项对主题相关性的作用、综合运用各类特征的通用化链接优先级计算公式以及特征选取原则。 By combing the various feature factors appearing in the search algorithm, a more comprehensive feature factor classification method is proposed, which divides the feature factors into six categories according to their own characteristics, namely the text content features, the Web topology features, the topic distribution features, the document Type features, statistical features and time-sensitive features, and describes the role of each feature on the relevance of the theme, comprehensive use of various types of common link priority calculation formula and feature selection principles.