本文選題：K均值 + 檢索結(jié)果聚類　； 參考：《計(jì)算機(jī)工程與設(shè)計(jì)》2017年04期

【摘要】：針對(duì)搜索引擎檢索大規(guī)模數(shù)據(jù)時(shí)結(jié)果聚類的性能有限問題,提出一種查詢?nèi)罩据o助的改進(jìn)K-Means算法。將傳統(tǒng)的K-Means聚類擴(kuò)展為多層次聚類的形式,實(shí)現(xiàn)檢索對(duì)象與檢索結(jié)果之間的聚類;通過引入檢索日志,輔助提升聚類的效果,實(shí)現(xiàn)檢索結(jié)果推送的高相關(guān)性。實(shí)現(xiàn)結(jié)果表明,基于該算法的檢索結(jié)果聚類,有著較高的準(zhǔn)確率,檢索過程的時(shí)間開銷較低,綜合效率與準(zhǔn)確率而言,該算法是一種理想的檢索結(jié)果聚類方法。
[Abstract]:Aiming at the limited performance of result clustering when searching large scale data by search engine, an improved K-Means algorithm with query log assistance is proposed. The traditional K-Means clustering is extended to multi-level clustering to realize the clustering between the retrieval objects and the retrieval results. The retrieval log is introduced to help improve the clustering effect and to achieve the high correlation of the retrieval results push. The results show that the algorithm has high accuracy and low time cost. The algorithm is an ideal clustering method for retrieval results in terms of both efficiency and accuracy.
【作者單位】： 江西科技師范大學(xué)黨委組織部;南昌理工學(xué)院經(jīng)濟(jì)管理學(xué)院;
【分類號(hào)】：TP391.3
，

本文編號(hào)：1787319