Automated News Summarization using Transformers
Preety Singh, Pranith Kashetty, Annaji Sai Rahul Teja Reddy, Gadeela Sai Teja, V. Sai Anusha · 2024
News summarization condenses newspaper articles for faster and easier consumption. Nowadays accessing news has become simple and easy. In this day and age, due to the rapid flow of information, numerous articles are as far as a touch of screen, but have limited time to consume the information, so it is required to create a news summarization system to access the summary of news and this task is done by using a simple T5 model, an abstractive transformer-based model, developed by Google AI. Pre-trained on the C4 dataset, the T5 model can be fine-tuned for the summarization process. The aim is to improve comprehension and efficiency when reading news articles. T5's abstractive approach enables it to generate summaries by understanding the context and meaning of the text. This allows the system to handle large news articles and extract the most important information. This system works on the transformer model called T5 Model, which allows the summarizing of the news articles by the process of configuration, tokenization, fine-tuning, encoding-decoding and evaluation. The system has conducted trials against some news articles and it has managed to summarize and through comparing the results with the human-generated summaries, a success rate of 55% was obtained.