We present a new parallel corpus, JHU FLuency-Extended GUG corpus (JFLEG) for developing and evaluating grammatical error correction (GEC). Unlike other corpora, it represents a broad range of language proficiency levels and uses holistic fluency edits to not only correct grammatical errors but also make the original text more native sounding. We describe the types of corrections made and benchmark four leading GEC systems on this corpus, identifying specific areas in which they do well and how they can improve. JFLEG fulfills the need for a new gold standard to properly assess the current state of GEC.
JFLEG: A Fluency Corpus and Benchmark for Grammatical Error Correction
The JHU FLuency-Extended GUG corpus (JFLEG) provides a comprehensive evaluation tool for grammatical error correction systems by addressing language proficiency levels and improving text fluency.
- Year
- 2017
- Venue
- jfleg-a-fluency-corpus-and-benchmark-for-1
- Authors
- 3
- Hosting
- Abstract onlyARXIV-DEFAULT
Cite
Notes
Only stored in your browser.
Attribution
- Abstract & full text
- arxiv.org/abs/1702.04066ARXIV-DEFAULT
- TL;DR
- Semantic Scholar