In Exploratory Data Analysis (EDA) for Natural Language Understanding (NLU), creating n-gram models is essential for understanding the contextual relationships between words, as highlighted in NVIDIA’s Generative AI and LLMs course. N-grams (e.g., bigrams, trigrams) capture sequences of words, revealing patterns and dependencies in text, such as common phrases or syntactic structures, which are critical for NLU tasks like text generation or classification. Unlike single-word frequency analysis, n-grams provide insight into how words relate to each other in context. Option A is incorrect, as computing word frequencies focuses on individual terms, missing contextual relationships. Option B is wrong, as sentiment analysis targets overall text sentiment, not word relationships. Option C is inaccurate, as word clouds visualize frequency, not contextual patterns. The course notes: “N-gram models are used in EDA for NLU to analyze word sequence patterns, such as bigrams and trigrams, to understand contextual relationships in textual data.”