CRISP-T (Cross Industry Standard Process for Triangulation) is a comprehensive Python framework for analyzing textual and numerical data using advanced NLP, machine learning, and statistical techniques. It is designed for researchers and practitioners working with mixed data research in fields like qualitative research, social sciences, and healthcare.

This wiki provides detailed documentation for CRISP-T version 2.0.

• Data Import Commands: Learn how to import your raw data (PDFs, TXT, CSV) into a CRISP-T corpus using crisp --source.

• Data Linking Commands: Link unstructured text documents to structured numeric data using IDs, keywords, time, or embeddings.
• Data Filtering Commands: Filter your corpus by metadata, time ranges, or relationships before analysis.
• Corpus Manipulation Commands: Advanced tools to manage, inspect, and manually edit your corpus structure using crispt.

• Data Analysis Commands: Perform text analysis (topics, sentiment, summaries) and numeric analysis (stats) using crisp.
• Machine Learning Commands: Run predictive models (regression, classification, LSTM) and clustering algorithms.
• Semantic Search Commands: Find documents by meaning, semantic chunks, similar documents, and more.
• Cross-Modal Analysis: Summarize cross-modal analysis capabilities, using text metadata as ML outcomes.

• Visualization Commands: Create publication-ready charts, word clouds, network graphs, and interactive plots with crispviz.

• Examples and Workflows: Step-by-step guides for common research scenarios (Mixed Methods, Qualitative Analysis).
• Utilities and Troubleshooting: Common flags, best practices, and solutions for common errors.

• Collaborative Sense-making with AI (MCP Server) — use MCP tools for assisted analysis

CRISP-T facilitates Methodological Triangulation by allowing you to seamlessly move between qualitative (text) and quantitative (numbers) data. By linking these modalities, you can validate findings from one method with evidence from another, strengthening your research conclusions.

Latest version: v2.0