Input should be two result files and a ground truth file.
The output should be a synopsis of differences of the output of the two results, per query .
For example, method B is better than A by +1% better MAP, because query xx gives a significantly better AP.
Input should be two result files and a ground truth file.
The output should be a synopsis of differences of the output of the two results, per query .
For example, method B is better than A by +1% better MAP, because query xx gives a significantly better AP.