Cluster-based visualization of user traffic on an internet site

A method that displays web traffic, comprising:

• receiving web site user clusters that were generated through one of zero-order and first-order Markov models, respective clusters including information related to one or more web pages accessed by one or more users who display similar web browsing characteristics;
• creating individual graphical user interfaces for each cluster;
• visualizing the cluster information within one or more rows of an associated graphical user interface;
• and
• sorting the one or more rows according to a probability of a behavior based on a location of a user within a cluster.

A method for displaying web site user activity according to web site behavior, comprising:

• utilizing an expectation-maximization algorithm to cluster web site users by browsing behavior;
• displaying user web site activity by visualizing information related to web pages accessed by users with similar behavior in fields displayed within a window;
• and
• employing different intensity levels with respective fields to indicate a probability of visiting a type of web page.

A computer readable medium storing computer executable components that facilitates visualizing web site activity, comprising:

• a component that partitions web site users according to similar web site navigation;
• a component that generates a window for each partition within a display;
• a component that displays information relating to web pages visited by the users in associated windows;
• and
• a component that sorts the web pages information within fields of a plurality of rows that respectively correspond to individual users, the fields displayed employing different intensity levels to indicate a probability of a user visiting a type of web page, the intensity levels ranging from low to high, a low intensity indicating a low probability of visiting the web page, and a high intensity indicating a high probability of visiting the web page.

A system that facilitates visualization web site activity, the system comprising program modules embodied on a computer-readable storage medium that facilitate:

• means for clustering web site activity information according to web site user behavior;
• means for displaying the clustered web site activity information in respective windows;
• means for presenting the web site activity information within respective windows based on similar user browsing behavior;
• and
• means for delineating the web site activity information into one or more rows within the respective windows, each row represents an individual user of the web site, the respective rows associated with an intensity level indicating a conditional transition probability of a zero-order or a first-order Markov model.