Step5.Normalize the hdv(i,j) for maxhdv(i,j)=1
Step6.Compute newdocument vectors
Step7. Goto Step4 until document vectorsconverge.
Step8. Computecumulated weight DCi(m) for each main concept cm for di

and normalize themfor max DCi(m)-1
Step9.Compute new 算法的第一步,我们扩展查询结果的根集来形成基本集,并且收集基本集中的所有超链接{l1,l2,l3,…..,lr},这里r是超链接l i= 2/2 首页 上一页 1 2 |