Sentence tokenization of sentences that do not have periods What is the current state of art solution for tokenizing text which lack period into sentences ? Is this even possible? if so, what program/implementation do you recommend for this task (R, python, Java,  etc)?
For example,
hi I am a boy I am a student I like food

into
hi.
I am a boy.
I am a student.
I like food.

 A: If you are looking specifically for a ML approach to the problem, you may want to check the RNN - in particular Character-Level Language Models. This article has explanation of them, and the author provides code on Github. The examples in the article manage to deal with code syntax, so the model should be fine for predicting punctuation, as long as you have a suitable training dataset.
However, ML perspective may not be the best approach, as it will probably retain some non zero error rate, regardless of the amount of training. I imagine that a tree parsing of a sentence can provide indication of transitioning from simple sentence to a complex one. This parser detects superfluous word(s) which could be used as a test whether the provided text should be still broken down into separate sentences.
A: Take any reasonable word-level RNN model (or character if truly necessary), download the Wikipedia corpus or really any large text corpus. Then form a training set of sentences where the RNN predicts whether there should be a period or not. In other words take a batch of sentences and remove he periods, and then have the RNN predict their locations. 
