Unifying Learning to Rank and Domain Adaptation: Enabling Cross-Task Document Scoring

For document scoring, although learning to rank and domain adaptation are treated as two different problems in previous works, we discover that they actually share the same challenge of adapting keywo