Dataset Similarity to Assess Semisupervised Learning Under Distribution Mismatch Between the Labeled and Unlabeled Datasets | Publicación