
✐
✐
“4137X˙CH08˙Akerkar” — 2007/9/14 — 15:28 — page 290 — #2
✐
✐
✐
✐
✐
✐
290 CHAPTER 8 Web Structure Mining
We will discuss some of the techniques that are useful in modeling web topology in subse-
quent sections.
8.2 Modeling Web Topology
We have seen that in information retrieval, we usually rank documents as a function of frequen-
cies of query terms within the document and across all documents. This method works very
well if most queries are long and well specified. Moreover, the documents in a single collection
such as CISI or MEDLAR are coherent, well authored, and are mostly about one topic. We
can also assume that the vocabulary contained in these do ...