Semantic – Enhanced Auto Encoder for Cyber Issues

Abstract

We propose another portrayal learning strategy to handle this issue. Our strategy named Semantic-Enhanced Marginalized Denoising Auto-Encoder (smSDA) is produced by means of semantic augmentation of the prominent profound learning model stacked denoising auto-encoder. The semantic augmentation comprises of semantic dropout clamor and sparsity limitations, where the semantic dropout commotion is composed in light of space learning and the word inserting strategy. Our proposed strategy can abuse the shrouded highlight structure of harassing data and take in a hearty and discriminative portrayal of content. Far reaching probes two open cyberbullying corpora (Twitter and MySpace) are led and the outcomes demonstrate that our proposed approaches outflank other benchmark content portrayal learning strategies.