Towards understanding word embeddings: Automatically explaining similarity of terms | IEEE Conference Publication | IEEE Xplore