I'm wondering if anyone knows of such natural language processing algorithm that can perform both of these functions. It seems from my searching that any algorithm either does word completion or next phrase completion but never both. The best way I can think to implement this is with dictionaries: unigram and n-gram. These have their cons tho and I'm hoping to implement something that is more state-of-the-art. The biggest challenge I've found with using NLP is the adapting to user input functionality, such that the suggestions are more tailored to what the user typically types. Whereas with tries you can up and down the frequencies of certain unigrams and n-grams as the user uses the autocomplete, NLPs are pretrained to be used and real time. I'm not the most well versed in natural language processing, so, overall, I would love some up-to-date pointers on resources and suggestions. Or, if this just doesn't exist yet, please let me know.
I have seen some other posts about this topic, but the NLP field has been a lot of advances in recent years since these posts so I wanted to renew this question.