Abstract:
In developing countries, Short Messaging Service (SMS) is
one of the most widely used and cheapest modes of commu-
nication. Hence, this medium is often exploited by advertis-
ing companies to reach masses. The unsolicited (spam) SM-
Ses consume user attention and have become a reason of an-
noyance for most of the mobile phone users, as not many of
them use the information from these SMSes. We conducted
a three phase study to understand the scale of SMS spam
problem and to propose technological measures to curb it.
First, we conducted a survey among 458 participants in In-
dia to understand general perceptions about spam SMSes,
user preferences and requirements. Ninety seven percent of
the survey participants admitted that they are quite annoyed
with the burst of spam SMSes and lack of appropriate tech-
nological or regulatory solutions.
Next, we designed and implemented a mobile based appli-
cation SMSAssassin, which can filter spam SMSes using
content based machine learning techniques and user gener-
ated customized rules. And last, we conducted a user study
of this application with twenty three participants who used
the application in real world for a month. Results show that
a mobile based solution can effectively filter spam SMSes
and the application can be usable too.