Rahul Garg and Naresh Kumar Garg, Maharaja Ranjit Singh Punjab Technical University, India

Text summarization in last few years have witnessed significant advancements and showcased various approaches and techniques. While there were notable strengths, it is important to consider the limitations and challenges that emerged during this period. One of the strengths observed in text summarization literature in last few years was the continued progress in deep learning models. Transformer-based architectures, such as BERT and GPT, continued to dominate the field and achieved impressive results in summarization tasks. These models effectively captured the contextual information and semantic relationships in the text, leading to more accurate and coherent summaries. However, there is much less work done in low resource languages like Punjabi due to their complex contextual structure and low availability of standard dataset. In this paper, we have presented a methodology that can be followed to generate summaries of Punjabi Text.

Keywords