Called SMSAssassin, the software is the brainchild of a team of researchers including professor Dr Vinayak Nai, assistant professor Dr. Ponnurangam Kumaraguru and a student Kuldeep Yadav.
The software, which was developed for Nokia’s Symbian operating system based phones, combines Bayesian filtering (a statistical technique of filtering spams in e-mail) with user submitted blacklists that uses ‘crowd sourcing’ to identify spam. The researchers are currently working on developing the software for Android based phones and those that run on Windows mobile.
SMS messages are very short, and often use regional terms or abbreviations. This makes it difficult to identify spam using traditional markers. According to Kumaraguru, the ‘crowd sourcing’ aspect enables the system to constantly update itself to new trends and changing tactics of spammers, and respond quickly to spam targeted to specific time periods, such as holidays.
According to the reports, despite a national DND registry, an estimated 100 million spam SMSes are sent to Indian mobile phone users every day.
This has given the researchers a plenty of source material to work with. With so many carriers, and no central place to parse these messages, the research team is banking on social media to help, build and train the system. “We are currently requesting the Facebook users to forward spam SMS messages to +91 8826068429.”
“SMSAssassin performs the spam filtering automatically. It is designed to run on mid-range phones wherein it filters such spam messages and also gives user freedom to receive SMSes which are spams but still useful to him or her,” Kumaraguru said.
“This technology is in a prototype stage and we hope to build a full-fledged system that can be deployed in real time soon,” said Kuldeep, the lead developer of SMSAssasin.