Understanding web pornography usage from traffic analysis

Pornography is massively available on the Internet, often free of charge. It represents a significant fraction overall Internet traffic, with thousands websites and millions users. Studying web pornography consumption useful to understand human behavior, it crucial for different disciplines, helping in sociological, statistical behavioral research. However, given lack public datasets, most works build surveys, limited by multiple factors, e.g., unreliable answers that volunteers may (even unconsciously) give. In this work, we analyze anonymized accesses using HTTP-level traces collected from an operational network. Our dataset includes traffic about 15000 broadband subscribers over three years. We use provide quantitative figures pornographic website consumption, focusing time frequency use, habits, trends. also compare users’ interests those who do not consume pornography, showing notable differences.