Intelligent Analysis of Textual Content for Spam Detection

Abstract

In this chapter, we presented several experiments that we conducted to evaluate the different models we proposed to detect spam in tweets. We introduced two different datasets and discussed how they were collected and how we used different sampling techniques to train and test the models. We presented the evaluation metrics we used and presented the results of each model in details. We compared our results to a previous study and showed that some of our models are outperforming their models and discussed the results we obtained from each model. At the end of the chapter, we have further discussed the results we obtained and made some observations that need more investigation. The character n-gram experiments presented in this chapter have been published in