Findings of the Association for Computational Linguistics: ACL 2023

By Zhiying Jiang et al

Published on July 9, 2014

Read the original document by opening this link in a new tab.

Table of Contents

1 Introduction
2 Related Work
3 Our Approach
4 Experimental Setup
5 Results

Summary

The paper presents a non-parametric alternative to DNNs for text classification using a combination of a simple compressor like gzip with a k-nearest-neighbor classifier. The method achieves competitive results with non-pretrained deep learning methods and even outperforms BERT on out-of-distribution datasets. It excels in the few-shot setting with scarce labeled data. The approach is simple, lightweight, and universal, offering a viable alternative to complex deep neural networks.