Automatic Text Summarization Using Graph-Based Recurrent Attention Model (GBRAM)

Text summarization is a system that compresses information to extract significant information from longer texts. It has grown to be a demanding area of study in natural language processing. Although the deep learning-based text summary model is currently producing good results, there is still room for improvement in terms of modeling word relationships, extracting feature information more precisely, and getting rid of redundant data. The modified graph-based recurrent attention model, which depends on recurrent attention, is proposed in this research. It significantly increases the readability and accuracy of text summarization. Recursive neural networks have a subclass called recurrent neural networks, which attempt to anticipate the following sequence by taking into account information from past states and the current state. A text data set was used for experimental validation, and the findings demonstrated that the model in this work performed better than previous approaches.