Research on NLP Based Automatic Summarization Generation Method for Medical Texts

Abstract

The fundamental concept underpinning text summarization technology revolves around the capacity to encapsulate the original information into a succinct form, thus equipping individuals to promptly extract essential content from vast data repositories and liberating users from the cumbersome task of processing extensive textual material. In recent years, the exponential proliferation of data in biomedical literature, patient case records, and healthcare documentation, has presented a pressing challenge. This research undertakes the integration of Natural Language Processing (NLP)-related technologies into the domain of medical text summarization. It puts forth a novel solution for generative automatic summarization, with a specific focus on enhancing the model's proficiency in assimilating the semantic nuances inherent in biomedical texts. The methodology incorporates within existing text summarization frameworks to optimize the model's efficacy in handling biomedical data. The empirical findings presented in this study attest to the remarkable precision of the sentence similarity calculation method introduced herein. In a comparative analysis against four alternative methodologies, this approach achieves a high accuracy rate of 90.6%. This outcome highlights the superior predictive performance of the sentence integration similarity calculation method proposed in this research.