Multimodal Fuzzy Framework Detects Fake Indian News Using Images and Text

The F2IND-IT! framework is a novel multimodal system designed to identify fake news within Indian media by integrating visual and textual analysis. It employs a ResNet-50 CNN for extracting image features, utilizes a DistilBERT encoder for generating text embeddings, and incorporates an Adaptive Neuro-Fuzzy Inference System (ANFIS) to yield a fuzzy reliability score. A lightweight attention-based fusion module is implemented to assign learnable weights to each modality prior to classification. When tested on the IFND dataset, this model surpasses earlier methods in terms of accuracy, precision, recall, and F1-score. This research tackles the issue of biased manipulation prevalent in both regional and national media outlets in India's varied media environment.