Sentiment analysis and random forest to classify LLM versus human source applied to Scientific Texts