Challenge Description

Text classification is a fundamental task in natural language processing, used to assign text to predefined categories. This challenge requires you to write a prompt that enables AI to accurately classify text into specified categories and provide a confidence score for the classification.

Challenge Goals

Write a prompt that enables AI to:

1. Accurately identify the topic and content features of the text.
2. Correctly classify the text into the specified categories.
3. Provide a confidence score for the classification (0-100%).
4. Explain the main basis for the classification.
5. Handle ambiguous or boundary case texts.

Requirements

• The classification result must include the category, confidence, and explanation.
• Confidence must be an integer from 0-100.
• The explanation must be concise (no more than 30 characters).
• For ambiguous text, select the most appropriate category and explain the uncertainty.
• The output format must be uniform and easy to parse.

Prompt Template

[Your Prompt]

Text: {Text Content}

Automated Testing Process

1. For each test case, use your prompt as input.
2. Analyze whether the output includes:
   • A clear classification result.
   • A confidence score from 0-100.
   • An explanation of no more than 30 characters.
3. Compare the classification result with expert-annotated standard answers.
4. Check the relevance of confidence to text clarity.
5. Evaluate the relevance and accuracy of the explanation.
6. Calculate the final score (out of 10 points).

Expected Output Example (for Case 1)

Classification Result:

  - Category: Technology
  - Confidence: 95%
  - Explanation: Involves Apple's new product release and technical specifications.

Reason: The text explicitly mentions Apple's new iPhone release, including technical details like the A17 chip and performance data, typical of technology news.

---

Submit your solution and share your ideas and techniques with the community!