What is the difference between classical NLP and NLP based on Deep Learning?

Classical NLP Modeling Output Stopword Removal(EN) Feature Extraction(EN) Modeling (English) Inference (English) Sentiment Pre-processing English Tokenization Pos Tagging (English) (English) Spanish Tokenization POS Tagging (Spanish) (Spanish) Arabic Tokenization POS Tagging (Arabic) (Arabic) Stopword Removal(ES) Feature i Extraction(ES) Modeling (Spanish) Inference (Spanish) Classification Feature Stopword Removal(AR) Extraction(AR) Modeling (Arabic) Inference (Arabic) Entity Extraction Translation Language Detection Topic Modelling -------- Documents Deep Learning-based NLP Output Sentiment Classification Preprocessing Entity Extraction Translation Topic Modelling Documents Hidden Layers Output Units Dense Embeddings obtained via word2vec, doc2vec, Glove, etc