Enhancing Question Answering Precision with Optimized Vector Retrieval and Instructions

TL;DR

This paper introduces an optimized vector retrieval method combined with instruction techniques to enhance question-answering accuracy, reducing computational costs and improving performance over traditional segmentation approaches.

Contribution

It presents a novel retrieval augmentation framework that optimizes document chunking and similarity functions to improve QA results without extensive fine-tuning of large language models.

Findings

Small chunk size (100 tokens) yields best QA performance.

Non-overlapping chunks outperform semantic sentence segmentation.

Optimized retrieval improves QA accuracy efficiently.

Abstract

Question-answering (QA) is an important application of Information Retrieval (IR) and language models, and the latest trend is toward pre-trained large neural networks with embedding parameters. Augmenting QA performances with these LLMs requires intensive computational resources for fine-tuning. We propose an innovative approach to improve QA task performances by integrating optimized vector retrievals and instruction methodologies. Based on retrieval augmentation, the process involves document embedding, vector retrieval, and context construction for optimal QA results. We experiment with different combinations of text segmentation techniques and similarity functions, and analyze their impacts on QA performances. Results show that the model with a small chunk size of 100 without any overlap of the chunks achieves the best result and outperforms the models based on semantic segmentation using sentences. We discuss related QA examples and offer insight into how model performances are improved within the two-stage framework.