1. Tokenize the text into words/phrases. 2. For each token, check if it's a proper noun (capitalized, not part of a proper noun phrase). 3. If not a proper noun, find three synonyms or variants that fit the grammatical context. 4. Replace with v1.
But how to handle multi-word proper nouns like "New York"? In the example, it's kept as is. So, when tokens are "New" and ManyVids 22 12 25 Andre Stone And Korra Del Rio...
First, I need to parse the example input. The example given is: "John is a student from France studying in New York. He loves to read books and play football." The user wants terms that aren't proper nouns to be replaced with three variants. Proper nouns here are John, France, New York, and football. Wait, no, football here is a term; since it's a common noun referring to the sport, maybe it's not a proper noun. But in the example, they replaced "study" with research, so it's about verbs, nouns, etc., not the proper nouns. For each token, check if it's a proper
Now, when applying this to a new input, I need to go through each word, determine if it's a proper noun (like John, France, New York), and if not, replace it with three variants. The example shows that even common nouns like "student" are replaced. when given a new input
Another challenge is handling contractions or hyphenated terms, but maybe the user expects each token to be considered individually.
So, when given a new input, the process is: