Back to Search Start Over

Native Language Identification with Big Bird Embeddings

Authors :
Kramp, Sergey
Cassani, Giovanni
Emmery, Chris
Publication Year :
2023

Abstract

Native Language Identification (NLI) intends to classify an author's native language based on their writing in another language. Historically, the task has heavily relied on time-consuming linguistic feature engineering, and transformer-based NLI models have thus far failed to offer effective, practical alternatives. The current work investigates if input size is a limiting factor, and shows that classifiers trained using Big Bird embeddings outperform linguistic feature engineering models by a large margin on the Reddit-L2 dataset. Additionally, we provide further insight into input length dependencies, show consistent out-of-sample performance, and qualitatively analyze the embedding space. Given the effectiveness and computational efficiency of this method, we believe it offers a promising avenue for future NLI work.

Details

Database :
arXiv
Publication Type :
Report
Accession number :
edsarx.2309.06923
Document Type :
Working Paper