figshare
Browse

Relevance-Based Extractive Summarization Framework for Document Management Systems

Download (207.52 kB)
software
posted on 2025-01-01, 03:48 authored by Krishna Chaitanya SunkaraKrishna Chaitanya Sunkara

This software presents a novel framework for automated document summarization, leveraging relevance scoring techniques to extract the most pertinent sections of a document. The solution was conceptualized and implemented in 2017, prior to the rise of generative AI technologies. The framework is designed to handle large documents efficiently, providing accurate, concise summaries for downstream applications. This approach enhances productivity in sectors such as legal, insurance, and accounting, where efficient information retrieval is essential. Unlike modern neural summarization models, Relevance employs a deterministic approach using relevance scoring, focusing on practical implementation and integration with existing DMS.

History