Efficient News Synopsis Generation with T5 Transformers

In this paper we propose a Text-To-Text Transfer Transformer-based (T5) model to generate concise, human-like synopses of Merger and Acquisition (M&A) economic news. Leveraging recent advancements in artificial intelligence and natural language processing, we have focused our resources on the creation of a specific summarizer, opting for a translation of all news into English to ensure consistency and mitigate complexity. The implementation of Graphics processing unit (GPU) resources for model training played a crucial role in optimizing the learning process, allowing us to effectively utilize the time window provided by platforms such as Google Colab Pro and Free. Through iterative fine-tuning, our model was trained to generate high-quality, human-like synopses of M&A events. Our findings underscore the enhanced transparency and interpretability of the proposed generative AI-driven text summariser, suggesting promising avenues for future exploration and development in this field.