With the binary model work going well, its time to move to the multi-label classification model and before we can annotate a large batch of article-repositories pairs, we need to nail down the annotation schema.
Going to annotate a sample of 100 article-repository pairs from our existing training and evaluation set.