NLP model and finetuning for journal

Viewed 11

I have a somewhat vague question, but maybe you can help me out. I have a kind of journal, where I've been writing an entry per day (just a few sentences), to describe my activities, for the past 5 years or so. I am wondering what cool things I could do with it, with a big pretrained NLP model. In particular, I'd be interested in:

  • Question answering: I would like to be able to ask questions such as "When did I last see X?" or "What places did I go to last year?"
  • Entry generation (for fun), maybe conditioned on week day, season, text input...

I'm guessing the first task might be a bit tricky, given the very specific format of the data, but the second one should be easy, right? How would you go about this? What other cool and useful tasks would you try if you had such data?

I work in research in computer vision, so I am familiar with using deep learning, and deep learning librairies. My knowledge of NLP is limited however. I am not afraid of delving a bit into the subject, and/or into some coding, but the easier it is to setup, the better.

Thank you for your help! If this is not the place for such questions, please let me know.

0 Answers
Related