Paraphrasing vs Coreferring: Two Sides of the Same Coin.

摘要: We study the potential synergy between two different NLP tasks, both confronting lexical variability: identifying predicate paraphrases and event coreference resolution. First, we used annotations from an dataset as distant supervision to re-score heuristically-extracted paraphrases. The new scoring gained more than 18 points in average precision upon their ranking by original method. Then, same re-ranking features additional inputs a state-of-the-art resolution model, which yielded modest but consistent improvements model's performance. results suggest promising direction leverage data models for each of tasks benefit other.