计算机工程
計算機工程
계산궤공정
COMPUTER ENGINEERING
2010年
2期
36-38
,共3页
网页搜索%相关性排序%编辑距离%字符串匹配
網頁搜索%相關性排序%編輯距離%字符串匹配
망혈수색%상관성배서%편집거리%자부천필배
Web search%relevance ranking%edit distance%string match
针对传统方法不能很好地处理网页中简短域和用户查询之间的相关性排序问题,提出一种改进的编辑距离(MED)排序算法,在编码和计算过程中引入查询词分布的位置、顺序和距离等信息,将查询和简短域之间的相关性问题转化为编码字符串的相似性问题.仿真实验结果表明,与传统的相关性排序算法相比,该算法可以提高网页搜索中简短网页域的相关性排序性能.
針對傳統方法不能很好地處理網頁中簡短域和用戶查詢之間的相關性排序問題,提齣一種改進的編輯距離(MED)排序算法,在編碼和計算過程中引入查詢詞分佈的位置、順序和距離等信息,將查詢和簡短域之間的相關性問題轉化為編碼字符串的相似性問題.倣真實驗結果錶明,與傳統的相關性排序算法相比,該算法可以提高網頁搜索中簡短網頁域的相關性排序性能.
침대전통방법불능흔호지처리망혈중간단역화용호사순지간적상관성배서문제,제출일충개진적편집거리(MED)배서산법,재편마화계산과정중인입사순사분포적위치、순서화거리등신식,장사순화간단역지간적상관성문제전화위편마자부천적상사성문제.방진실험결과표명,여전통적상관성배서산법상비,해산법가이제고망혈수색중간단망혈역적상관성배서성능.
Aiming at the problems that the traditional methods can not perform well on the short Web page fields, a Modified Edit Distance(MED) algorithm is proposed. In the process of encoding and calculating, the algorithm uses the position, order, and distance information, so the problem on the relevance between the corresponding query and short field can be converted to the problem on the similarity between the encoding strings. Simulation experimental results show this algorithm can significantly outperform the traditional algorithms for relevance ranking on short Web fields, especially for very short fields.
