# Learning Paraphrastic Sentence Embeddings from Back-Translated Bitext

## TL;DR

This paper introduces a method for learning paraphrastic sentence embeddings using back-translation of bilingual data, achieving high-quality training data comparable to manual paraphrases, scalable across languages and domains.

## Contribution

The authors propose leveraging neural machine translation for generating large-scale paraphrase datasets, improving data quality and scalability for training sentence embeddings.

## Key findings

- Back-translation produces high-quality paraphrase data.
- Data quality rivals manually-created paraphrases.
- Neural machine translation outputs differ from human sentences in length and vocabulary.

## Abstract

We consider the problem of learning general-purpose, paraphrastic sentence embeddings in the setting of Wieting et al. (2016b). We use neural machine translation to generate sentential paraphrases via back-translation of bilingual sentence pairs. We evaluate the paraphrase pairs by their ability to serve as training data for learning paraphrastic sentence embeddings. We find that the data quality is stronger than prior work based on bitext and on par with manually-written English paraphrase pairs, with the advantage that our approach can scale up to generate large training sets for many languages and domains. We experiment with several language pairs and data sources, and develop a variety of data filtering techniques. In the process, we explore how neural machine translation output differs from human-written sentences, finding clear differences in length, the amount of repetition, and the use of rare words.