Semantic technologies for detecting names of new drugs on darknets

Abstract

There is an emerging international phenomenon of drugs that are sold without any control on online marketplaces. An example of a former online marketplace is Silk Road, best known as a platform for selling illegal drugs operated as a Tor hidden service. Silk Road was closed by FBI in 2013 but new alternatives have appeared since illicit substances is a big market. One problem with online marketplaces is that the sold substances have many different names and new substances are constantly developed. In this work we use semantic techniques to automatically detect new names of drugs. Our experiments are applied on data from a darknet marketplace, on which we use a set of known drug names and distributional statistics to find words that are semantically similar. The results show that semantic technologies work very well when it comes to detecting names of drugs on darknets.