Video spam comment features selection using machine learning techniques

Abstract

<p>Nowadays, social media (e.g., YouTube and Facebook) provides connection and interaction between people by posting comments or videos. In fact, comments are a part of contents in a website that can attract spammer to spreading phishing, malware or advertising. Due to existing malicious users that can spread malware or phishing in the comments, this work proposes a technique used for video sharing spam comments feature detection. The first phase of the methodology used in this work is dataset collection. For this experiment, a dataset from UCI Machine Learning repository is used. In the next phase, the development of framework and experimentation. The dataset will be pre-processed using tokenization and lemmatization process. After that, the features to detect spam is selected and the experiments for classification were performed by using six classifiers which are Random Tree, Random Forest, Naïve Bayes, KStar, Decision Table, and Decision Stump. The result shows the highest accuracy is 90.57% and the lowest was 58.86%.</p>

Analysis of spammers’ behavior on a live streaming chat

“…Each message was manually labeled as being relevant or irrelevant. In the experiment, evaluation of the classification of user behavior was analyzed with 10-fold cross-validation using 6 different types of classifiers [1] that is naïve bayes [15,[22][23][24], decision trees [15,24], random forest [15,25], k-nearest neighbors [26], support vector machine (SVM) [27][28], and artificial neural network (ANN) [29]. The result of this classification was analyzed to find suitable features for identifying spam during live streaming.…”

“…Spam is an unwanted message that consists of texts and links. Besides insulting posts, mass messaging, cruelty, humiliation, hate speech, malicious, fake hints [1], or fraudulent reviews, most spams messages are intended for advertising purposes: explicit advertising message or links to malicious websites in the form of long URLs or short URLs (Google URL Shortener, Bitly, TinyURL) redirecting to a product's website in order to increase the site's rating or promote products.…”

BotDetectorFW: an optimized botnet detection framework based on five features-distance measures supported by comparisons of four machine learning classifiers using CICIDS2017 dataset

“…Extra features can cause an increase in the computation time; also can add negative impact to an accuracy of the detection system. Therefore; it is advisable to reduce the number of the features in a dataset using a suitable feature selection technique that searches for the best set of features that could optimize a classification of data [14]. In literature, feature selection techniques are classed into filter [15], wrapper [16], and embedded techniques [17].…”

Corona fault detection in switchgear with extreme learning machine

“…It is simple in theory and fast in implementation. Literature study indicates that the ELM has significantly higher learning speed compared to that of a traditional feed-forward network learning algorithms while showing better generalization performance [17][18][19][20]. Based on empirical risk minimization theory, the learning process of the ELM requires only a single iteration.…”