Due to the continual growth of the popularity of the Internet, commercial a
s well as industrial companies have been advertising their products and ser
vices via the Web, resulting in a drastic increase in the number of Web sit
es. With a huge amount of information available on various Web sites, it is
important that the relevant and useful information favored by individual v
isitors is delivered to the destinations in a timely manner. The two tradit
ional approaches for sorting web information including search engines and h
ierarchical indices require specific input by the visitors who may not have
any specific favorite sites in mind. In most cases, site surfers are just
"window-shopping" on the Internet, looking for "exciting" things. This pape
r proposes the development of an Intelligent Internet Information Delivery
System (IIIDS) which is characterized by its machine learning capability ba
sed on the data of site spots "movements" by the users within the Web pages
and then evaluates the site preferences of the relevant users by means of
fuzzy logic principle. The development of IIIDS and the test of a prototype
to evaluate its feasibility are covered in this paper. (C) 2000 Elsevier S
cience Ltd. All rights reserved.