Coder Social home page Coder Social logo

arag_eval's Introduction

ARAG: Analysis and Retrieval Augmented Generation for Comprehensive Reasoning over Socioeconomic Data (Performance Evaluation)

Abstract

Recent advancements in Large Language Models (LLMs) have significantly impacted the field of question answering systems, particularly with LLM-based data analysis and Retrieval-Augmented Generation (RAG). Yet, their independent application has limited the scope in scenarios requiring a synthesis of both data analysis and contemporary information retrieval. To bridge this gap, we introduce the Analysis and Retrieval Augmented Generation (ARAG) framework, which blends data analysis with the retrieval of up-to-date information. Based on the framework, we build a system for interpreting the dynamics of socioeconomic indicators and how ARAG explores the reasons behind the indicators' dynamics through data analysis of correlated indicators and retrieval of relevant facts from news sources. The comparison of ARAG with Microsoft Copilot and Perplexity showed that ARAG significantly outperformed in delivering in-depth analytical insights and was remarkably more robust to misinformation across various queries.

Results

Chart 1
Comparison of ARAG with Microsoft Copilot and Perplexity.
Chart 2
Ablation study of ARAG.

Data

  • arag_results.xlsx: Responses of ARAG, MS Copilot, and Perplexity over 50 queries (Chinese).
  • arag_results_en.xlsx: Responses of ARAG, MS Copilot, and Perplexity in English traslated via ERNIE Bot 4.0.
  • negative_sample_results_arag.xlsx: Responses of ARAG over 20 queries with misleading information.
  • negative_sample_results_copilot.xlsx: Responses of MS Copilot over 20 queries with misleading information.
  • negative_sample_results_perplexity.xlsx: Responses of Perplexity over 20 queries with misleading information.

Result Files

  • arag_results.xlsx: Evaluation results ARAG, MS Copilot, and Perplexity over 50 queries and 10 realizations.
  • average_scores.csv: Average socres of ARAG, MS Copilot, and Perplexity over 50 queries.

arag_eval's People

Contributors

ethan-yxx avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.