external/FinGPT/fingpt/FinGPT_RAG/multisource_retrieval/README.md

FinGPT Sentiment Analysis

Motivations:

• Classify financial statements to help traders aggregate and digest financial news

Methods:

• FinGPT fine-tuning
• Retrieval-Augmented Generation

Setup

• Set up your .env file, can refer to /FinGPT_sentiment/.env.example
• Visit FinGPT_RAG/requirements.txt for the environment setup

python news_scraper.py

I. Data Preparation

Task 1: GPT-based News Classification

1. On UI, select csv file to load
2. Creates "classification" column for each financial statement
3. Using "default_classification_prompt" to ask GPT to classify the news
4. Saves .csv

Task 2: Context Retrieval

1. On UI, select csv file to load
2. Creates "contextualized_sentence" for each financial statement
3. Using Google and various news sources to retrieve the context
   4. Add relevant news paragraphs to form "contextualized_sentence" for each financial statement
4. Saves .csv

Experiment:compare non-RAG vs. RAG sentiment classification using gpt:

1. utils/sentiment_classification_by_external_LLMs.py: Call openAI APIs to classify RAG-based and non-RAG-based statements
2. utils/classification_accuracy_verification.py: Calculate the accuracy of sentiment classification between RAG-based and non-RAG-based statements
3. Results: 0.7876588021778584 vs. 0.8130671506352088

python utils/sentiment_classification_by_external_LLMs.py
python utils/classification_accuracy_verification.py