text-as-data-social

What problem does it solve?

This Skill solves the problem of converting unstructured social science text into quantitative, analysis-ready variables without relying on ad-hoc NLP guesswork.

Core Features & Use Cases

• Dictionary-based content analysis: Build and apply LIWC-style dictionaries to generate interpretable category proportions and counts (e.g., frames, ideology-like word categories).
• Topic modeling at scale: Discover themes with LDA or BERTopic and use embeddings to make topics meaningful and comparable across documents.
• Sentiment and semantic measurement: Compute sentiment scores and perform embedding-based similarity and (optionally) bias-related measures.
• LLM-assisted annotation support: Use LLM labeling workflows with validation principles to translate qualitative coding schemes into datasets that can be modeled.

Quick Start

Ask your Claude agent to process your corpus, run BERTopic topic modeling, and output a document-by-topic feature table plus sentiment scores.