As the title says, any pointers are much appreciated.
I am exploring, where do we stand in terms of ML/NLP efforts, in context of solving (to begin with - parsing) Math Word Problems.
We have decent enough softwares in likes of Mathematica which can solve well formulated math equations.
But when it comes to solving math problems expressed in natural languages, I could not find anything substantial. When I think about how to approach this, I see it as a sort of Machine Translation problem (translating from English to Math-equations), but there is hardly any 'labeled' data for that. Other approach can be semi (or un) sypervised Relation Extraction.
Since these are just random thoughts, I want to start with some existing work/papers in this direction. My otherwise decent googling skills, didn't help much.
[link][5 comments]