Enhancing Spam Detection Accuracy using Genetic Algorithm

The technology advancement has resulted in increased usage of internet which in turn has provided certain groups a chance to misuse the resources and services. Due to this exploitation, spam messages are sent to the customers because of which they are missing out on the important information. Spam has become a critical problem in online social networks. Spam in social networks refers to the unwanted, malicious, unsolicited content or behavior, fake friends, fraudulent reviews. Algorithms used for spam detection can take a lot of time to be trained as it includes irrelevant and redundant features. This results in poor predictions and high computational overhead. The proposed method mainly focuses on selecting the best features to increase the accuracy. Thus, selecting relevant feature subsets can help in reducing the computational cost and speed up the learning process. An extensive research was done to implement Apriori technique as well as genetic algorithm on different email datasets, along with feature extraction and pre-processing. The comparison of our results show the best suitable model is also discussed