🏗️Taxonomy Builder

Purpose: The Taxonomy Builder workflow is designed to create a structured taxonomy for organizing a project’s data and content. This taxonomy can be used across various agentic workflows to enhance searchability, topic categorization, and overall content organization. The goal is to establish a clear, logical content structure that other workflows can leverage to improve functionality.

Prerequisites:

• A comprehensive set of project documentation.

• Taxonomy Agent configured and ready to process.

• Loader Agents (e.g., GitBook, GitHub, Website) to ingest content that will be categorized.

How it Works:

1. The Loader Agents collect project documentation and content.

2. The Taxonomy Agent analyzes the content and generates a hierarchical taxonomy based on the data structure.

3. The resulting taxonomy can be applied to workflows like topic classification, improving content discovery and organization.

Inputs:

• Project documentation and content from sources like GitBook, GitHub, or websites.

• Configurations or guidelines for the desired taxonomy structure.

Processing:

• The Taxonomy Agent scans and processes the input content, applying logic to generate a meaningful taxonomy.

• It organizes content into categories and subcategories based on predefined rules or machine learning models.

Outputs:

• A structured taxonomy, often presented as a hierarchy or cluster of topics and categories.

• The taxonomy can be exported or integrated into other workflows for use.

Usage: Admins initiate the Taxonomy Builder workflow through the admin panel, selecting content sources and defining taxonomy rules.

Example: The Taxonomy Agent organizes a software development project’s documentation into categories like "Installation," "API Documentation," and "Error Handling," with subcategories for detailed topics.

Advanced Features:

• Customizable rules for how the taxonomy is built, such as adding weight to certain keywords or topics.

Troubleshooting:

• Misclassified Content: Adjust the taxonomy rules or retrain the model for better accuracy.

Logs:

• Logs can be accessed under Taxonomy Logs, showing details on how the content was categorized.

Related Workflows:

• Topic Classification Agent for applying the taxonomy to real-time data.

• Content Improvement Workflow for identifying gaps based on the taxonomy.