DERI Online Browsing Behavior Study


In a nutshell: The idea behind DOBBS is to get a deeper understanding how users browse the Web (e.g., how much time they spent on the Web, how long they stay on a website, how often they visit a website, how they use their browser, etc.). To contribute, DOBBS provides a Firefox browser add-on that keeps track of users’ browsing behavior. DOBBS addresses privacy concerns in a straightforward fashion: no personal data are collected, and all sensitive data (i.e. URLs of visited web pages) are encrypted before being sent to the server.


What DOBBS can do

To give an idea what DOBBS will be able to provide, we show some first, straightforward results for a very small dataset consisting of three users.

(1) Two important aspects of DOBBS are how long users are idling and how long a Firefox window is not the active window on the desktop. The first figure shows the average for both figures for all three users. User 1, for example, was most of the time idling but the Firefox window had always the focus. This indicates that User 1 mostly leaves the computer running with Firefox open and in the foreground. For User 2 and 3, on the other hand, the idle time and time the window is in the foreground are very similar. This means that Firefox was actually in the background when the users were not using it.

(2) The second figure shows for all three users the absolute number of sessions and the average number of page loads per session. User 1 obviously prefers short, individual sessions – in most cases, “session” refers to the time from opening and closing a Firefox window. User 3, on the other hand, is using the same Firefox window for a long time span – DOBBS, of course, allows to determin the length of these time spans – for browsing the Web. User 2 lies in between these “extreme” behaviors.

(3) Multi-tabbing is supported by all modern browsers, and is a frequently used feature. As a result, loaded web pages are not necessarily on display. DOBBS enables to investigate how long pages are actually in the foreground, compared to the time they are loaded in th browser. The third figure shows these numbers for a single browsing session of one user. The upper figure shows the times for each individual page load; the lower one shows the aggregated values for each domain. Note that in most cases the time a page is loaded but not on display is relatively long.

(4) The detailed records of DOBBS allow to reconstruct the exact browsing history – in terms of activities – of users. The graph in the forth figure shows such a history for an individual session. Nodes represent page loads with the size of a node indicating the time the pages was loaded in a tab. Edges represent the navigation between web pages. Nodes with more then one out-going edge indicate the use of multiple tabs. Such a representation allows the application of all kinds of graph-based measures (diameter, longest path, etc.) to describe, quantify, and categorize the browsing behavior of users


The idea behind DOBBS

Originally, browsing the Web has mainly been considered the active task of searching for information. With today’s available bandwidth resources and the resulting success of new kinds of online platforms, this has changed significantly. Particularly media or streaming sites, like YouTube or, invite users to watch video clips or listen to radio in a rather passive fashion. Further, many web pages are dynamic, periodically updating their content without explicit user interaction.

The huge success of social media sites such as Facebook or Twitter also had a significant effect on users’ browsing behavior. Firstly, they attracted a whole new audience of users that previously was less inclined to frequently use the Web, if at all. And secondly, as studies show, social media sites have strong sociological impact indicated by the increasing time users spend online and how they arrange their social life accordingly.

Modern browsers also allow performing active and passive browsing activities in parallel. For example, a user can search for information while listing to an online radio station in a background tab and occasionally check the latest – automatically updated – stock data in second browser window. As a result, browsing the Web now covers a very broad spectrum of activities. It even does no longer involve that users are sitting in front of their PC or laptop during browsing session.

All these developments give way to new and unexplored forms of online browsing behavior. This is where DOBBS comes in. DOBBS is aimed to get a thorough understanding how users browse the Web. For this, DOBBS provides an easy to install browser add-on – available for Firefox – that keeps track of users’ browsing behavior in a completely anonymized manner. To be more specific, the add-on observes:

  • how often and long users visit web pages
  • how people navigate between pages (e.g. by clicking on links or bookmarks)
  • how many browser windows or tabs user during a session
  • how often users switch between multiple tabs
  • how often and long users are idling, i.e. not actively using the browser window


Benefits of DOBBS

Having deeper insights into the browsing behavior of users allows deriving both fundamental and applied knowledge from different perspectives

  • The information how long and “how active” a user visits a website provides implicit feedback about its quality in nature, potentially allowing for more sophisticated ranking algorithms of Web search engines. For example, a site that typically resides a background tab might be considered differently than sites that typically involve more active user participation.
  • With the browser as major application to access the Web, knowledge about users’ browsing behavior enables the design and development of new features that improve the online experience of users. Examples include the hiding of and the quick access to tabs containing passively used web pages (e.g. online radio), or the automatic rearranging of tabs according to their usage.
  • From a more technical perspective, new optimization approaches to save bandwidth or computing resources are conceivable. This might include, for example, special “idle modes” for browsers where pages are not automatically updated.
  • And finally, understanding how and when user browse the Web also allows to derive statements on why they use it (e.g., for information seeking, entertainment, socializing, etc.). This, in turn, provides further insights into the sociological impact of the Web — that is, how the Web and “being online” shapes peoples’ life.


People behind DOBBS

DOBBS is an initiative of the Digital Enterprise Institute of Technology (DERI) of the National University of Ireland, Galway (NUIG). Directly involved in this project are:

Dr. Christian von der Weth
Prof. Manfred Hauswirth

Leave a Reply

Your email address will not be published.

4 + = thirteen

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>