Research idea - mapping haskell type signatures to real world questions
Hoogle is sort of what I'm thinking about. But Hoggle requires well-typed info to find the functions you are looking for. And hoogle does not string together functions into solutions or programs.
In #haskell, all day long, there is question after question such as "How can I get a list of all pairs from this list" or "how do I apply a function to the first element of a list and then take that result and apply it to the next element of the list?"
Can we take the next step from hoogle and take english language requests and turn that into haskell code?
Can we build a browser for Haskell functions that allows one to navigate a tree of semantic concepts?
Most importantly, can we get inside the head of a Haskell programmer and break down his solutions into easy steps to make it easier for others to learn haskell?