A pattern matching model for identifying deception

Abstract

Social network is a platform among several people that are bounded together on the basis of some relationships. Deception during communication on social network is very common and this has increased fraud and criminal activities on the social media network. Hence, this study builds a pattern matching model that identifies and curbs deceptive behaviors on social media network, using twitter as a case study. Conversational trust algorithm was used to establish a relationship between two people on the social network while proportionality algorithm was used to analyze the messages between them. The proportionality algorithm used potential propagations which satisfy a constraint. It makes use of words stored in an array to cross validate the conversation to detect deceptive words. Brute force algorithm, which is a pattern matching algorithm, was used to detect deceptive words and the algorithms were implemented using PHP. The results showed that the developed model is capable of establishing a relationship and identifying deception based on the frequency of the appearance of the deceptive words.

Keywords: Social network, Deception, Conversational trust algorithm, Proportionality algorithm, Brute force algorithm.