Information Technology Reference
In-Depth Information
REGFORM2: here the visitor has to subscribe a contract in which he/she accepts
the conditions for on-line commerce HELP: it answers questions that may arise
during the navigation through the web site;
FDBACK: a page that allows to go back to the previous one visited
FDPOST: a page that allows to go back to one previously seen page, in determined
areas of the site
NEWS: it presents the last novelties available;
SHELF: it contains the list of the programs that can be downloaded from the
website
PROGRAM: gives detailed information on the characteristics of the software
programs that can be bought;
PROMO: gives an example (demo) of the peculiarities of a certain program;
DOWNLOAD: it allows to download software programs of interest;
CATALOG: it contains a complete list of the products on sale in the web site;
PRODUCT: shows detailed information on each product that can be purchased;
P_INFO: a page on which detailed information on the terms of payment of the
products can be found;
ADDCART: the place where the virtual basket can be filled with items to be
purchased;
CART: shows the current status of the basket, that is, which items it contains;
MDFYCART: allows to modify the current content of the basket, for instance
taking off items;
CHARGE : indicates the amount due to buy the items contained in the basket;
PAY_REQ: a page which visualizes the amount finally due for the products in the
basket;
PAY_RES: here the visitor agrees to pay, and data for payment are inserted (for
example, the credit card number);
FREEZE: where the requested payment can be suspended, for instance to add new
products to the basket.
3 Exploratory Data Analysis
Our main aim is to discover the most frequent sequence rules among the 36 binary
variables describing whether any single page has been visited. In order to obtain valid
conclusions, the considered data have to be homogeneous. In order to assess whether
the visitors dataset (and, correspondingly, the sequence dataset that generates it) is
homogeneous we have deemed necessary to run first an exploratory analysis on the
visitors dataset. The available quantitative variables, which are somehow ancillary for
the analysis, can be usefully employed for this purpose.
On the basis of the explanatory analysis, we have decided to proceed with outliers
removal. We have decided to eliminate all observations above the 99-th percentile of
