Internet Browsing History Data Analysis for Automatic Negative Content Website Identification (Case Study: TRUST+™ Positif)

Negative content website is a website that contains one or more of these following elements: pornography, violence and coercion in children, incitement to anarchy, and gambling. Negative content website grows along with the development of the internet. The number of internet users who's potentially exposed to negative content is increasing because of the cheaper cost of access to the internet and the increasing number of devices that support the use of the internet. Several programs have been taken by the authorities. Among them is the creation of TRUST + ™ Positive system that holds the huge list of negative website address. The ISP (Internet Service Provider) will blocks any negative content website referring to this system. The number of negative content website listed in the TRUST + ™ Positive list increases when there's reports about new negative content website or after doing the back-crawling process. The problem we faced is that the addition of the number of negative content website listed on the TRUST + ™ Positive list depends heavily on external reports and the ability of the TRUST + ™ Positive back-crawling engine. Therefore, by using ISP's Internet browsing history data we will performing data mining process to identify new negative content website. Data mining is done by using an association algorithm. Some internet user browsing history data setup techniques are used to find the best results according to internet browsing patterns that may arise. To reduce the number of identification errors we will filter any websites that are believed to be a website that has no negative content. The result obtained is that although the results of the association algorithm can be used for the identification of negative content website, but more than 75% of those results are not a negative content website or still need validation about its content.