NLP techniques applied to web analytics

Repository of our paper entitled "Natural Language Processing for Web Browsing Analytics: Challenges, Lessons Learned, and Opportunities" submitted to Computer Networks

Authors: Daniel Perdices, Javier Ramos, José Luis García-Dorado, Iván González and Jorge E. López de Vergara.

Contents of the repository

• dataset/ contains the datasets analyzed in the paper.

  • top100/ has samples obtained with five diferents User-Agents (U1, ..., U5) of the top 100 domains of Alexa.
  • top2500/ has samples of the top 2500 domains of Alexa.

  Filenames follow the next syntax:

  <date>_<browser>.txt
  

  and their content is the analysis of the DNS packets. In particular, each line is:

  <timestamp> <counter of the number of responses in DNS packet> <question> <answer (only different from question in CNAME responses)> <TTL>
  

• examples/ contains some examples of the methods. Soon to appear.