Hotlink enhancement algorithms for web directories (extended abstract)

Consider a web site containing a collection of web pages with data. Each page is associated with a weight representing the frequency that page is accessed by users. In the tree hierarchy representation, accessing each page requires the user to travel along the path leading to it from the root. By enhancing the index tree with additional edges (hotlinks) one may reduce the access cost of the system. That is, the hotlinks reduce the expected number of steps needed to reach a leaf page from the tree root, assuming that the user knows which hotlinks to take. The hotlink enhancement problem involves finding a set of hotlinks minimizing this cost. The paper proposes a hotlinks structure allowing a user with limited a-priori knowledge to determine which hotlink to use at every given point. Then, a polynomial algorithm is presented for solving the hotlink enhancement problem for such hotlinks on trees of logarithmic depth. The solution is first presented for binary trees and then extended to arbitrary degree trees. It is also shown how to generalize the solution to situations where more than one hotlink per node is allowed. The case in which the distribution on the leaves is unknown is discussed as well, and is given an algorithm guaranteeing (an optimal) logarithmic upper bound on the expected number of steps down the tree.