西南师范大学学报(自然科学版)
西南師範大學學報(自然科學版)
서남사범대학학보(자연과학판)
Journal of Southwest China Normal University (Natural Science Edition)
2015年
9期
78-84
,共7页
李红波%孟欣赏%吴渝%李娜芬
李紅波%孟訢賞%吳渝%李娜芬
리홍파%맹흔상%오투%리나분
数据挖掘%匿名用户%用户识别%浏览路径%访问时长
數據挖掘%匿名用戶%用戶識彆%瀏覽路徑%訪問時長
수거알굴%닉명용호%용호식별%류람로경%방문시장
data mining%anonymous user%user identification%access path%brow sing time
有效的用户识别与用户细分是网站用户行为分析的基础。针对现有用户识别算法将注册用户和匿名用户均按匿名用户处理,导致用户分类不细致的问题,提出了一种匿名用户识别算法。该算法通过识别用户访问行为状态,采取页面访问路径和浏览时长匹配方式,进一步识别IP地址变化后混入纯匿名用户中的注册匿名用户,从而把用户细分为注册用户、假匿名用户和纯匿名用户。实验结果表明,该算法能够提高匿名用户识别率,更加准确地识别假匿名用户。
有效的用戶識彆與用戶細分是網站用戶行為分析的基礎。針對現有用戶識彆算法將註冊用戶和匿名用戶均按匿名用戶處理,導緻用戶分類不細緻的問題,提齣瞭一種匿名用戶識彆算法。該算法通過識彆用戶訪問行為狀態,採取頁麵訪問路徑和瀏覽時長匹配方式,進一步識彆IP地阯變化後混入純匿名用戶中的註冊匿名用戶,從而把用戶細分為註冊用戶、假匿名用戶和純匿名用戶。實驗結果錶明,該算法能夠提高匿名用戶識彆率,更加準確地識彆假匿名用戶。
유효적용호식별여용호세분시망참용호행위분석적기출。침대현유용호식별산법장주책용호화닉명용호균안닉명용호처리,도치용호분류불세치적문제,제출료일충닉명용호식별산법。해산법통과식별용호방문행위상태,채취혈면방문로경화류람시장필배방식,진일보식별IP지지변화후혼입순닉명용호중적주책닉명용호,종이파용호세분위주책용호、가닉명용호화순닉명용호。실험결과표명,해산법능구제고닉명용호식별솔,경가준학지식별가닉명용호。
Valid user identification and user segment are the basis of user behavior analysis .For the user classification is oversimplified w hich leads by confusing anonymous users and registered users all as anony‐mous user in existing user identification algorithms ,an anonymous user identification algorithm has been proposed .By identifying users'access behavior state and matching users'access path and brow sing time length ,the algorithm identifies the registered anonymous user mixed in the pure anonymous user due to the IP address change to classify user into registered login user ,registered anonymous user ,pure anony‐mous users .The experimental results indicate that the proposed algorithm can improve anonymous user i‐dentification rate ,and identify the registered anonymous user much more accurately .