USING LLM DOCUMENT CLASSIFICATION AT LOCAL DISK

Abstract

Using LLM Document classification at Local disk, Automated document classification is the machine learning 
fundamental that refers to assigning automatic categories among scanned images and files of the documents. It 
reached the state-of-art stage but it needs to verify the performance and efficiency of the algorithm by comparing. 
The objective was to get the most efficient classification algorithms according to the usage of the fundamentals 
of LLM. This project focuses on the development of an automated document categorization system for a local 
disk, leveraging a Large Language Model (LLM) and zero-shot classification techniques. The primary goal is to 
classify and organize documents based on both their content and file extension, automatically moving them to 
their corresponding folders. Users can either download the documents or pass them to the application through a 
command line or API, after which the system identifies the document's extension, analyses its content using a pre
trained LLM, renames the file based on its contents, and then moves it to the appropriate folder. 
Keywords: Document classification; machine learning algorithms; LLM; Zero Short Technique; analysis