# ParaAMR: A Large-Scale Syntactically Diverse Paraphrase Dataset by AMR Back-Translation

Source: [ParaAMR: A Large-Scale Syntactically Diverse Paraphrase Dataset by AMR Back-Translation](<https://www.sota2.com/research/paper/68fa6f801e209131cea46051>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-02T02:39:26.742Z

Authors: Kuan-Hao Huang, Varun Iyer, I-Hung Hsu, Anoop Kumar, Kai-Wei Chang, Aram Galstyan

Published: 2023-05-26T02:27:33.000Z

## Abstract

Paraphrase generation is a long-standing task in natural language processing (NLP). Supervised paraphrase generation models, which rely on human-annotated paraphrase pairs, are cost-inefficient and hard to scale up. On the other hand, automatically annotated paraphrase pairs (e.g., by machine back-translation), usually suffer from the lack of syntactic diversity -- the generated paraphrase sentences are very similar to the source sentences in terms of syntax. In this work, we present ParaAMR, a large-scale syntactically diverse paraphrase dataset created by abstract meaning representation back-translation. Our quantitative analysis, qualitative examples, and human evaluation demonstrate that the paraphrases of ParaAMR are syntactically more diverse compared to existing large-scale paraphrase datasets while preserving good semantic similarity. In addition, we show that ParaAMR can be used to improve on three NLP tasks: learning sentence embeddings, syntactically controlled paraphrase generation, and data augmentation for few-shot learning. Our results thus showcase the potential of ParaAMR for improving various NLP applications.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Paraphrase Syntactic Diversity on Human Evaluation 300 samples (random sample)](<https://www.sota2.com/research/sota/paraphrase-syntactic-diversity-on-human-evaluation-300-samples-random-sample.md>): Paraphrase Syntactic Diversity; Human Evaluation 300 samples (random sample); Score 3 (%): 0.182.

- [Paraphrase Similarity and Diversity Evaluation on Shared subset of 193,869 source sentences (PARANMT, PARABANK, PARAAMR)](<https://www.sota2.com/research/sota/paraphrase-similarity-and-diversity-evaluation-on-shared-subset-of-193-869-source-sentences-paranmt-parabank-paraamr.md>): Paraphrase Similarity and Diversity Evaluation; Shared subset of 193,869 source sentences (PARANMT, PARABANK, PARAAMR); Semantic Similarity: 82.05.

- [Paraphrase Semantic Similarity on Human Evaluation 300 samples (random sample)](<https://www.sota2.com/research/sota/paraphrase-semantic-similarity-on-human-evaluation-300-samples-random-sample.md>): Paraphrase Semantic Similarity; Human Evaluation 300 samples (random sample); Score 3: 26.5.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2305.16585>)

[PDF](<https://arxiv.org/pdf/2305.16585.pdf>)

[Implementation](<https://github.com/uclanlp/ParaAMR>)