The distance calculation of inner product is wrong.

[CaucherWang]

The inner product of two vectors represents the similarity. So we should return the largest ones as kNN. It's opposite to L2 distance.

[CongFu92]

Currently we only support Euclidean distance. Nearest neighbor search mainly refers to Euclidean space. Maximum inner product problem is another active literature in info retrieval. We will consider scale our work to it in future