Our Analysis Process
Political Speech Input
User provides political speech text in plain format.
Text Preprocessing
Tokenize and clean input. Break down into individual statements.
Fact/Propaganda Segmentation
Utilize LLMs (e.g., Gemini, DeepSeek) to classify statements as FACTUAL or PROPAGANDISTIC.
Intent & Emotional Tone Detection
For propagandistic segments, identify emotional tone (e.g., fear, hope) and intent (e.g., influence market).
Stock Market Reaction Mapping
Map factual/propaganda elements to historical market responses. Estimate % change per sector (tech, pharma, etc.).
Final Output Generation
Generate a comprehensive report including statements, market impact, intent summary, overall forecast, and confidence score.
Speech Analysis
Enter a political speech below. The AI will analyze it for factual and propagandistic statements, emotional tone, intent, and predict potential market impact.