Then get a team together and enter the Hewlett Foundation's contest focusing on automated scoring of student essays. The goal is to accelerate innovation in computer-based essay scoring to make complex questions more feasible for state tests. High costs and delays in scoring essay questions make most states opt only for multiple-choice questions.
Both the PARCC and Smarter Balanced assessment consortia are supporting the competition.
The competition will proceed in two phases. The first phase will focus on existing vendors and platforms for computer-based essay scoring, to characterize the state-of-the-art for consideration in the development of the actual competition.
The second phase, which opens later this spring and runs through April, will be open to the public and is a competition to see who can develop algorithms which most approximate human scoring results.
If this is a nut that can really be cracked, the implications (and applications) are enormous. And there's nothing like a big fat check to spur innovation...