Natural Language Processing

knowledge skill

Parses unstructured text to extract entities, sentiment, and semantic relationships. It produces structured data representations of linguistic patterns and intent.

Worked examples

• {"input": {"text": "Apple Inc. shares rose 2% in Cupertino today.", "task": "ner", "entities": ["ORG", "GPE"]}, "output": {"extracted_entities": [{"text": "Apple Inc.", "label": "ORG"}, {"text": "Cupertino", "label": "GPE"}]}}
• {"input": {"text": "The battery life is disappointing, but the screen is vivid.", "task": "sentiment"}, "output": {"sentiment_score": -0.2, "summary": "Mixed review focusing on battery and display."}}

Input

• text: string: raw text to analyze
• task: string: specific operation (e.g., 'ner', 'sentiment', 'summarization')
• entities: array: list of target labels for extraction

Output

• extracted_entities: array: list of identified terms and labels
• sentiment_score: float: polarity value from -1.0 to 1.0
• summary: string: condensed version of input text

Details

• Skill type: knowledge skill
• Safety level: safe_public_research
• Version: 1.0.0