Hello everyone,
Over the past few days I have been running several tests on the URL filtering system, specifically using the DBL lists and those developed by the University of Toulouse.
My goal was to perform some practical evaluations to better understand which of the two solutions could be more suitable for my use case, assessing aspects such as effectiveness, category coverage, and overall behavior in filtering web traffic.
From the tests I have carried out so far, both lists seem to behave quite similarly. In terms of general functionality, they both perform their task properly, successfully identifying and blocking a significant portion of domains that fall within their respective categories. In other words, from a purely operational perspective, the filtering mechanism appears stable and consistent with expectations.
However, during these tests I noticed something that I think is worth highlighting, particularly regarding the DBL lists. My impression is that they currently present some gaps in their categorization structure. To explain this more clearly: while analyzing the available categories, I noticed that some sections that would be quite relevant in a modern and comprehensive filtering environment appear to be missing.
A clear example concerns social networks. In many scenarios—especially in corporate, educational, or testing environments—it can be very useful to have a dedicated category that allows administrators to easily block platforms such as Facebook and similar services through centralized policies. Similarly, it would also be beneficial to have more specific categories dedicated to websites involved in the sale or promotion of drugs, weapons, or other illegal goods, enabling more granular and targeted filtering policies.
I fully understand that building and maintaining domain classification lists is a demanding and time-consuming task. It requires constant monitoring, frequent updates, and a significant amount of resources to ensure both accuracy and broad coverage. That said, in its current state, I feel that the DBL lists are somewhat limited in terms of category structure, which may reduce their effectiveness in certain operational contexts.
It would be very interesting to know whether there are plans to expand or refine the taxonomy of categories in the future, because with a few targeted improvements these lists could become an even more comprehensive and competitive tool.
If anyone has conducted similar tests or has experience with other filtering lists, I would be very interested in hearing your feedback and comparing different approaches.