Sentiment Analysis with Mistral 7B Instruct using Amazon Bedrock

Perform sentiment analysis on your data using Mistral 7B Instruct via Amazon Bedrock

This set of Data Productivity Cloud pipelines performs sentiment analysis on a set of product review data using the Mistral 7B Instruct Large Language Model (LLM).

How to run the pipelines

Start with Sentiment Analysis - Mistral 7B Instruct, as shown in the screenshot above. It does two things:

• Runs Sentiment Analysis - Acquire Review Data to create a set of product review data in a database table named stg_sample_reviews
• Run the actual sentiment analysis using the LLM, creating a new database table named stg_sample_reviews_genai

Once it has finished, open Sentiment Analysis - Comparison, as shown in the screenshot below.

Use Sample Data while the Comparison component is highlighted, to check how the AI-generated sentiment score matches the original numeric star rating of the review.

Downloads

Licensed under: Matillion Free Subscription License

• Download Sentiment-Analysis-Mistral-7B-Instruct-Amazon-Bedrock-Databricks.zip
  • Target: Databricks
• Download Sentiment-Analysis-Mistral-7B-Instruct-Amazon-Bedrock-Redshift.zip
  • Target: Redshift
• Download Sentiment-Analysis-Mistral-7B-Instruct-Amazon-Bedrock-Snowflake.zip
  • Target: Snowflake

Installation Instructions