Detecting Dravidian Offensive Posts in MIoT: A Hybrid Deep Learning Framework

Hate speech and Offensive Posts (OP) detection on Smart Multimedia Internet of Things (MIoT) have been an active issue for researchers. MIoT media texts in non-native English-speaking countries are often code-mixed or script mixed/switched. This paper proposes ensemble-based Deep Learning (DL) framework comprised a Convolutional Neural Network (CNN) Dense (DNN) identifying hate OP Malayalam Code-Mixed (MCM), Tamil (TCM), Script-Mixed (MSM) postings. Word-level character-level features utilized the convolutional neural network. In contrast, dense network uses Term Frequency-Inverse Document Frequency (TF-IDF) features. The inclusion proposed ensemble resulted state-of-the-art performance TCM MCM datasets, with weighted F 1 -score 0.91 0.78, respectively, comparable MSM posts, 0.95.