I'm looking to solve the following problem: I have a list of similar sentences as my dataset, and I want to be able to type a new sentence, which is also similar to the sentences in my dataset and then highlight the parts of the sentence which are varying. For example,
This is the set I have:
I want to eat an apple I want to eat an ice cream I want to drink coffee I want to eat a pizza I want to puke ginger
These sentences are similar but they have some varying parts. The two main variables in these sentences are the
action (eat, drink, puke) and the
food (apple, ice cream, coffee, pizza, ginger).
Now if I type in a new sentence say
I want to drink cola I want my model to identify
cola as the variables in this sentence.
At this point, I don't really care about the semantics that much. I will work on that later. Even if the sentence is
I want to hug a table, all I want to do is identify
table as variables.