Web Analytics Understanding user behavior and ... - pace university
Web Analytics Understanding user behavior and ... - pace university
Web Analytics Understanding user behavior and ... - pace university
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
The Problem<br />
Server log files from CSIS website are used in this thesis. The server log files (412 MB) are stored<br />
in the hard drive. The log files are then queried to extract only the logs where the <strong>user</strong>s used<br />
different keywords to reach the same web site.<br />
For Example:<br />
Keywords: <strong>pace</strong>, <strong>pace</strong> <strong>university</strong>, csis, graduate center of <strong>pace</strong> <strong>university</strong>, Westchester county<br />
courses<br />
Search Engines: AltaVista, yahoo.com, google.com etc.<br />
The <strong>Web</strong> page reached: / ( csis.<strong>pace</strong>.edu)<br />
The keywords <strong>and</strong> its relevance to the web site reached is understood manually by the human eye<br />
using knowledge <strong>and</strong> intuition.<br />
When the <strong>user</strong> reaches the web site, the activity recorded tells if the <strong>user</strong> ever wanted to be on the<br />
site. Was he conducting any business or came to the site by mistake. These conclusions are drawn<br />
by underst<strong>and</strong>ing the log files of each <strong>and</strong> every particular <strong>user</strong>. The activity of each <strong>user</strong> in the<br />
main data files is again mined form the source file by using 'grep'.<br />
The log file of the particular <strong>user</strong>, who has used a keyword in a search engine to reach this<br />
particular web site, is obtained. From this it is understood whether the <strong>user</strong> was serious in his<br />
search, or not.<br />
The room for error in such a prediction exists. Never the less, it is more important to know the<br />
bigger picture, so that a specific description, idea is obtained form these server log files about large<br />
number of <strong>user</strong>s.<br />
41