The rapid spread of misinformation on social media highlights the need for robust, automated fact correction frameworks.However, existing works rely on supervised learning from manually annotated claim-evidence pairs, which are scarce and prone to biases, limiting their generalization across domains.Moreover, these methods overlook semantic faithfulness in their correction process.To address these challenges, we propose Mask-to-Correct (M 2 C), a training-free, inference-only Retrieval Augmented Generation (RAG) based framework that leverages diversity-aware masking to identify erroneous spans of claims and evaluate the faithfulness of corrections using retrieved evidence.However, the effectiveness of RAG heavily depends on the choice of retriever, which may vary across queries.To mitigate this, we further introduce M 2 C + , an ensemblebased framework that combines corrections across multiple rankers to reduce retrieval bias and improve robustness.Extensive experiments on the benchmark datasets demonstrate that our proposed frameworks consistently outperform all baselines, achieving up to 14% improvement in SARI scores, without using gold evidence.

Leveraging both labeled (input-output associations) and unlabeled data (wider contextual grounding) may provide complementary benefits in retrieval augmented generation (RAG). However, effectively combining evidence from these heterogeneous sources is challenging as the respective similarity scores are not inter-comparable. Additionally, aggregating beliefs from the outputs of multiple rankers can improve the effectiveness of RAG. Our proposed method first aggregates the top-documents from a number of IR models using a standard rank fusion technique for each source (labeled and unlabeled). Next, we standardize the retrieval score distributions within each source by applying z-score transformation before merging the top-retrieved documents from the two sources. We evaluate our approach on the fact verification task, demonstrating that it consistently improves over the best-performing individual ranker or source and also shows better out-of-domain generalization.

Recent success in language generation capabilities of large language models (LLMs), such as GPT, Llama, etc., can potentially lead to concern about their possible misuse in inducing mass agitation and communal hatred via generating fake news and spreading misinformation. Traditional means of developing a misinformation ground-truth dataset do not scale well because of the extensive manual effort required to annotate the data. It is crucial to anticipate and counteract potential adversarial fake information to mitigate detrimental effects and promote societal harmony. To this end, this PhD proposal spans three main research directions. The first concerns investigating ways of developing unsupervised models for fake news identification leveraging retrieval augmented generation (RAG) approaches. In our second thread of work, we explore ways of creating synthetic datasets to eventually train supervised or few-shot example-based models. Another direction of research work involves tracking the propagation of fake information through social networks to develop preventive measures against it.