An Error-Oriented Approach to Word Embedding Pre-Training

TL;DR

This paper introduces an error-oriented pre-training method for word embeddings that leverages writing errors in learner scripts, outperforming traditional models and benefiting from error correction and increased data.

Contribution

The paper presents a novel approach that uses writing errors in learner scripts for pre-training word embeddings, improving performance over existing methods.

Findings

Error-oriented embeddings outperform traditional models.

Augmenting with error corrections enhances performance.

More data further improves results.

Abstract

We propose a novel word embedding pre-training approach that exploits writing errors in learners' scripts. We compare our method to previous models that tune the embeddings based on script scores and the discrimination between correct and corrupt word contexts in addition to the generic commonly-used embeddings pre-trained on large corpora. The comparison is achieved by using the aforementioned models to bootstrap a neural network that learns to predict a holistic score for scripts. Furthermore, we investigate augmenting our model with error corrections and monitor the impact on performance. Our results show that our error-oriented approach outperforms other comparable ones which is further demonstrated when training on more data. Additionally, extending the model with corrections provides further performance gains when data sparsity is an issue.