The widespread propagation of misinformation on social media platforms poses a significant concern, prompting substantial endeavors within the research community to develop robust detection solutions. Individuals often place unwavering trust in social networks, often without discerning the origins and authenticity of the information disseminated through these platforms. Hence, the identification of media-rich fake news necessitates an approach that adeptly leverages multimedia elements and effectively enhances detection accuracy. The ever-changing nature of cyberspace highlights the need for measures that may effectively resist the spread of media-rich fake news while protecting the integrity of information systems. This study introduces a robust approach for fake news detection, utilizing three publicly available datasets: WELFake, FakeNewsNet, and FakeNewsPrediction. We integrated FastText word embeddings with various Machine Learning and Deep Learning methods, further refining these algorithms with regularization and hyperparameter optimization to mitigate overfitting and promote model generalization. Notably, a hybrid model combining Convolutional Neural Networks and Long Short-Term Memory, enriched with FastText embeddings, surpassed other techniques in classification performance across all datasets, registering accuracy and F1-scores of 0.99, 0.97, and 0.99, respectively. Additionally, we utilized state-of-the-art transformer-based models such as BERT, XLNet, and RoBERTa, enhancing them through hyperparameter adjustments. These transformer models, surpassing traditional RNN-based frameworks, excel in managing syntactic nuances, thus aiding in semantic interpretation. In the concluding phase, explainable AI modeling was employed using Local Interpretable Model-Agnostic Explanations, and Latent Dirichlet Allocation to gain deeper insights into the model’s decision-making process.

The rise of social media has enabled individuals with biased perspectives to spread hate speech, directing it toward individuals based on characteristics such as race, gender, religion, or sexual orientation. Constructive interactions in varied communities can greatly enhance self-esteem, yet it is vital to consider that adverse comments may affect individuals’ social standing and emotional health. The crucial task of detecting and addressing this type of content is imperative for reducing its negative effects on communities and individuals alike. The rising occurrence highlights the urgency for enhanced methods and robust regulations on digital platforms to protect humans from such prejudicial and damaging conduct. Hate speech typically appears as a deliberate hostile action aimed at a particular group, often with the intent to demean or isolate them based on various facets of their identity. Research on hate speech predominantly targets resource-aware languages like English, German, and Chinese. Conversely, resource-limited languages, including European languages such as Italian, Spanish, and Portuguese, alongside Asian languages like Roman Urdu, Korean, and Indonesian, present obstacles. These challenges arise from a lack of linguistic resources, making the extraction of information a more strenuous task. This study is focused on the detection and improvement of multilingual hate speech detection across 13 different languages. To conduct a thorough analysis, we carried out a series of experiments that ranged from classical machine learning techniques and mainstream deep learning approaches to recent transformer-based methods. Through hyperparameter tuning, optimization techniques, and generative configurations, we achieved robust and generalized performance capable of effectively identifying hate speech across various dialects. Specifically, we achieved a notable enhancement in detection performance, with precision and recall metrics exceeding baseline models by up to 10% across several lesser-studied languages. Additionally, our work extends the capabilities of explainable AI within this context, offering deeper insights into model decisions, which is crucial for regulatory and ethical considerations in AI deployment. Our study presents substantial performance improvements across various datasets and languages through meticulous comparisons. For example, our model significantly outperformed existing benchmarks: it achieved F1-scores of 0.90 in German (GermEval-2018), up from the baseline score of 0.72, and 0.93 in German (GermEval-2021), a substantial increase from 0.58. Additionally, it scored 0.95 in Roman Urdu HS, surpassing the previous peak of 0.91. Furthermore, for mixed-language datasets such as Italian and English (AMI 2018), our accuracy rose dramatically from 0.59 to 0.96. These outcomes emphasize the robustness and versatility of our model, establishing a new standard for hate speech detection systems across diverse linguistic settings.

The rise of social media has led to an increase in hate speech. Hate speech is generally described as a deliberate act of aggression aimed at a particular group, intended to harm or marginalize them based on specific attributes of their identity. While positive interactions in diverse communities can greatly enhance confidence, it is important to acknowledge that negative remarks such as hate speech can weaken community unity and present a significant impact on people’s well-being. This highlights the need for improved monitoring and guidelines on social media platforms to protect individuals from discriminatory and harmful actions. Despite extensive research on resource-rich languages, such as English and German, the detection and analysis of hate speech in less-resourced languages, such as Norwegian, remains underexplored. Addressing this gap, our study leverages a metalinguistic approach that uses advanced meta-learning techniques to enhance the detection capabilities across bilingual texts, effectively linking technical advancements directly to the pressing social issue of hate speech. In this study, we introduce techniques that adapt models that deal with hate speech detection within the same languages (intra-lingual), across different languages (cross-lingual), and techniques that adapt models to new languages with minimal extra training, independent of the model type (cross-lingual model-agnostic meta-learning-based approaches) for bilingual text analysis in Norwegian and English. Our methodology incorporates components that help the model focus on relevant parts of the text (attention mechanisms) and tools that adjust the learning speed based on performance (adaptive learning rate schedulers). We conducted various experiments using language-specific and multilingual transformers. Among these, the combination of Nor-BERT and LSTM with zero-shot and few-shot model-agnostic meta-learning achieved remarkable F1 scores of 79% and 90%, highlighting the effectiveness of our proposed framework.