Unsupervised paraphrase acquisition has been an active research field in recent years, but its effective coverage and performance have rarely been evaluated. We propose a generic paraphrase-based approach for Relation Extraction (RE), aiming at a dual goal: obtaining an applicative evaluation scheme for paraphrase acquisition and obtaining a generic and largely unsupervised configuration for RE. We analyze the potential of our approach and evaluate an implemented prototype of it using an RE dataset. Our findings reveal a high potential for unsupervised paraphrase acquisition. We also identify the need for novel robust models for matching paraphrases in texts, which should address syntactic complexity and variability.