Refining a Polarity Classification Prompt
A data analyst is using a language model to classify customer feedback. The model is producing inconsistent results for reviews that contain both positive and negative comments. Analyze the provided prompt and example, identify the core issue causing this inconsistency, and describe a specific modification to the prompt that would improve its reliability for such cases.
0
1
Tags
Ch.3 Prompting - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Analysis in Bloom's Taxonomy
Cognitive Psychology
Psychology
Social Science
Empirical Science
Science
Related
Example Definition of the 'Positive' Category for Polarity Classification
Definition of the 'Negative' Category in Polarity Classification
A developer is using a language model to sort user feedback into three categories: 'Bug Report', 'Feature Request', and 'Praise'. The initial prompt is:
Classify the following feedback: [feedback text]. The developer observes that the model often misclassifies suggestions for improvement (e.g., 'It would be great if this button did X') as bug reports. Which of the following prompt modifications is most likely to resolve this specific ambiguity between the 'Bug Report' and 'Feature Request' categories?Definition of the 'Positive' Category in Polarity Classification
Refining a Polarity Classification Prompt
Improving a News Headline Classifier