Abstract:
The fast development of social networking
sites such as imparting, sharing, putting away and
overseeing huge data leads to pull in cybercriminals.
Spammers misuse these social networking sites to
abuse cyber laws for their unlawful arts. They start
with email, and then quickly spread to new
advancements, for example, texting, newsgroups and
smart phones. As online social networks, for
example, MySpace, Facebook and Twitter turned out
to be progressively well known, spammers rapidly
found another home for their spamming purposes.
Spamming activities of social spammers not only
causes dangerous for normal social network users
but also annoys to these users. The aim of this paper
is to develop social spammer detection approach with
low cost and low overhead. The detection approach
is a three-phase process: (1) features extraction, (2)
features selection and (3) classification. Validation of
this approach is tested with 1KS-10KN dataset and
CRESCI-2015 dataset.
