Steps

• Collect data (e.g. in crisp_source folder). Data includes multiple text documents (txt or pdf) and ONE numeric CSV file with (optional) timestamps.
• Import data into CRISP-T corpus and dataframe.
• Perform linking between text and numeric data using various methods (id based, keyword based, time based, embedding based).
• Explore text data using various methods (e.g., topic modeling, keyword extraction, sentiment analysis, visualizations).
• Explore numeric data using various methods (e.g., summary statistics, classification, clustering, regression, association, visualizations, TDA, etc.).
• Perform cross modal analysis using linked text and numeric data (e.g., text features as predictors for numeric outcomes, numeric features as predictors for text outcomes, etc.).
• Add manual connections between text documents and numeric rows if needed to support theory driven analysis.
• Derive insights from the analysis and document them.
• Use an AI agent to help with analysis, interpretation, and documentation if needed using MCP tools.