Introduction

A Dify plugin that integrates the Dingo data quality evaluation library to help automatically detect data quality issues in datasets and text content.

Features

• Text Quality Evaluation: Assess the quality of text content using built-in rules from the Dingo library
• Multiple Rule Groups: Support for different evaluation rule sets (default, sft, rag, hallucination, pretrain)
• Quality Scoring: Get numerical quality scores (0-100%) with detailed issue reports
• Simple Integration: Easy to use within Dify workflows, chatflows, and agent applications
• Local Processing: All evaluation happens locally, no external API calls required

Usage

1. Install the plugin in your Dify workspace
2. Add the "Text Quality Evaluator" tool to your workflow
3. Configure the tool parameters:
   • Text Content: The text you want to evaluate
   • Rule Group: Choose between "default", "sft", "rag", "hallucination", or "pretrain" rule sets
4. Get comprehensive quality assessment results including:
   • Overall quality score percentage
   • Number of issues detected
   • Detailed list of specific problems found

Example Use Cases

• Content Moderation: Evaluate user-generated content for quality issues
• Data Preprocessing: Clean datasets before training or analysis
• RAG System Enhancement: Improve retrieval quality by filtering low-quality documents
• Content Creation: Validate generated text meets quality standards

Rule Groups

The plugin supports different rule groups optimized for specific use cases:

Group	Use Case	Description
default	General text quality	Basic quality checks including content completeness, formatting issues
sft	Fine-tuning datasets	Rules from default plus hallucination detection for supervised fine-tuning
rag	RAG system evaluation	Response consistency and context alignment assessment
hallucination	Hallucination detection	Specialized rules for detecting AI-generated content issues
pretrain	Pre-training datasets	Comprehensive set of 20+ rules for large-scale dataset evaluation

About Dingo

Dingo is a comprehensive data quality evaluation tool that helps you automatically detect data quality issues in your datasets. Dingo provides a variety of built-in rules and model evaluation methods, and also supports custom evaluation methods. It supports commonly used text datasets and multimodal datasets, including pre-training datasets, fine-tuning datasets, and evaluation datasets.

Key Features of Dingo:

• Multi-source & Multi-modal Support: Local files, Hugging Face datasets, S3 storage
• Rule-based & Model-based Evaluation: 20+ built-in rules, LLM integration, hallucination detection
• Comprehensive Reporting: 7-dimensional quality assessment with detailed traceability

Contact & Support

• Plugin Repository: dingo-plugin
• Original Dingo Project: DataEval/dingo
• Issues & Feedback: Please report issues on the plugin repository
• Discord: Join Dingo Community
• Online Demo: Try Dingo on Hugging Face

License

This project uses the Apache 2.0 Open Source License.

Privacy Policy

This plugin processes text data locally within your Dify environment. No data is transmitted to external servers. See privacy-policy.md for full details.