Hearsay: Suppressing spam using trust in mobile social gossiping networks

The development of social networks has changed the way we communicate from a personal conversation into a broadcast with a world-wide audience. Social networks have proven themselves useful during protests and natural disasters by providing a platform to share ideas and offer help to those in need. However, due to the reliance on the Internet there have been situations where social networks were unable to function. Examples are broken submarine cables and countries that have restricted Internet access for residents, such as Egypt and Libya during the uprising in the beginning of 2011. Alternative communication methods are available, such as mobile ad-hoc networking, but suffer a common problem, large quantities unwanted messages, better known as spam. We propose a novel approach called "Hearsay" to combat spam. Our approach suppresses spam in gossiping networks by utilising social network information. Gossiping networks are characterised by users carrying devices capable of creating ad-hoc communication links with devices nearby. Messages are exchanged using a gossip protocol, spreading them throughout the network. Without suppression a message would reach every user making it ideal for spamming. Prioritising messages based on personal and social relations keeps messages within a social group. Users assign a rank to each user they interact with and based on that rank messages are forwarded to other members. Through simulations we show the feasibility of Hearsay, as a large-scale deployment is not possible within the set time frame. We ran our simulations within the multi-agent modeling environment NetLogo, which showed the effectiveness of using trust to suppress spam.

Subject

spam
pocket switched networks
trustrank
social networks
mobile
trust
ad-hoc networking
gossiping

To reference this document use:

http://resolver.tudelft.nl/uuid:2b257abd-a17a-43c0-938b-2b25fcead09a

Part of collection

Student theses

Document type

master thesis

Rights

Files

PDF

thesis.pdf

923.61 KB

Close viewer