Quelle: Transkripte des Weißen Hauses und der Kampagne (Januar 2025 – April 2026).
Werkzeuge: Python (Scrapy zur Datenerfassung), NLTK/SpaCy zur Verarbeitung natürlicher Sprache und Tokenisierung.
Von firehmre
Quelle: Transkripte des Weißen Hauses und der Kampagne (Januar 2025 – April 2026).
Werkzeuge: Python (Scrapy zur Datenerfassung), NLTK/SpaCy zur Verarbeitung natürlicher Sprache und Tokenisierung.
Von firehmre
7 Kommentare
The word „DEAL“ emerged as a statistically significant outlier
The term „nobody“ exhibits a higher weights-per-sentence ratio than most geopolitical entities, including „Canada“ and „Mexico.“
Except China is spelled wrong. When Trump says it, it’s Jina.
So…you removed a huge part of his vocabulary. Got it. Downvoted.
It might be improved by using word stems to count. For example “tariff” vs “tariffs” and “country” vs “countries”.
It would be improved by using word stems to count. For example “tariff” vs “tariffs” and “country” vs “countries”.
1. You should have included cuss words as well. This way you are „manipulating“ data because of you own beliefs
2. Good work, keep going
Why the fuck would you remove cuss words from this? Would be more interesting to see how many f bombs and shits were dropped. Why self censor words spoken by the president…weak.