Abstract
Social media in the last decade has become a popular communication mechanism on the web. Sites like Facebook, Twitter and YouTube are seeing enormous growth. It is important to understand the trends of this new type of media for many reasons including identity theft, social engineering, advertising and digital preservation. Some data sets have been made available to the public such as the tweets from Twitter, alternately data can be scraped from the open web. However, to ascertain trends from a group of individuals such as employees of a business, or students of a university, there is no way, without asking each individual
member, what social media sites they use. Within this paper, we present a detailed approach to gaining this type of information. Specically, for a group of geographically and organizationally aliated members, we present an unsupervised approach that can discover and disambiguate social media problems with a precision of 0.863 and an F-measure of 0.654.
Users
Please
log in to take part in the discussion (add own reviews or comments).