Why is smoothing so important for n-gram modeling? Please select all that apply. Note that partial credit is allowed but incorrect answers could get negative scores. A. Smoothing is the process of flattening a probability distribution implied by a language model so that all reasonable word sequences can occur with some probability. B. All smoothing methods "steal from the rich to give to the poor", however, add-one smoothing steals way too much. C. If an N-gram is never observed in the training data, it cannot occur in the evaluation data set. D. Smoothing is an intuitively simple concept. It pretends each n-gram occurs once more than it actually does in the training data set. E. Since MLE underfits the training data, smoothing methods reassign probability mass from observed to unobserved events to avoid overfitting/zero probabilities