I want to create an AI that received any unspaced book and is able to space at least 80% of it correctly. Is it even possible? If yes, In which AI technics should I use to make this work?
I need a database of unspaced books and their correct form so my AI can practice on them, is there is any database like so that you can link me to?
Thank you very much
Yes, you can make it with AI. There is a subfield in AI, which is called Natural Language Processing (NLP). It has a lot of algorithms for text specific datasets. You need to find a solution with NLP, which is indeed possible.
About the dataset, you can find it everywhere. Just take whatever text you want (e-book, Wikipedia's text data, etc..) and remove the spaces. You can collect a really big dataset for it.