To prepare 1 million sentences for Word2Vec embeddings (example: "The quck BROWN FOX jumps over the lazy dog"), which operations should be applied to sanitize and prepare the text consistently? (Choose three.)

To prepare text for Word2Vec, consistent preprocessing is crucial. Converting all characters to lowercase (normalization) ensures that "The" and "the" are treated as the same word, reducing vocabulary size and improving embedding quality. Tokenizing each sentence into individual words is fundamental, as Word2Vec operates on words, not full sentences. Removing stop words (common words like "a," "is," "the") helps Word2Vec focus on more meaningful content words, as stop words often add noise without significant semantic contribution. Part-of-speech tagging and retaining only verbs/nouns would drastically reduce context, which Word2Vec relies on to learn word relationships. Fixing spelling errors is beneficial but not a mandatory or primary step for basic Word2Vec preparation and can be computationally expensive. One-hot encoding every word is an alternative representation, not a preprocessing step for Word2Vec, which learns dense embeddings.