0 citations0 references

TSDAE: Using Transformer-based Sequential Denoising Auto-Encoderfor Unsupervised Sentence Embedding Learning

2021pp. 671–688

Citations Over TimeTop 10% of 2021 papers

Kexin Wang, Nils Reimers, Iryna Gurevych

Abstract

Learning sentence embeddings often requires a large amount of labeled data. However, for most tasks and domains, labeled data is seldom available and creating it is expensive. In this work, we present a new state-of-theart unsupervised method based on pre-trained Transformers and Sequential Denoising Auto-Encoder (TSDAE) which outperforms previous approaches by up to 6.4 points. It can achieve up to 93.1% of the performance of indomain supervised approaches. Further, we show that TSDAE is a strong domain adaptation and pre-training method for sentence embeddings, significantly outperforming other approaches like Masked Language Model. 1 A crucial shortcoming of previous studies is the narrow evaluation: Most work mainly evaluates on the single task of Semantic Textual Similarity (STS), which does not require any domain knowledge. It is unclear if these proposed methods generalize to other domains and tasks. We fill this gap and evaluate TS-DAE and other recent approaches on four different datasets from heterogeneous domains.

Related Papers

→ Embedding as a modeling problem(1998)167 cited
Design and Realization of RS Encoder Based on FPGA(2009)
→ A(2023)