# Binary Token-Level Classification with DeBERTa for All-Type MWE Identification: A Lightweight Approach with Linguistic Enhancement

Source: [Binary Token-Level Classification with DeBERTa for All-Type MWE Identification: A Lightweight Approach with Linguistic Enhancement](<https://www.sota2.com/research/paper/69799810a2b28e761d87cbe8>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-29T05:00:05.136Z

Authors: Diego Rossini, Lonneke van der Plas

Published: 2026-01-27T08:42:54.000Z

## Abstract

We present a comprehensive approach for multiword expression (MWE) identification that combines binary token-level classification, linguistic feature integration, and data augmentation. Our DeBERTa-v3-large model achieves 69.8% F1 on the CoAM dataset, surpassing the best results (Qwen-72B, 57.8% F1) on this dataset by 12 points while using 165x fewer parameters. We achieve this performance by (1) reformulating detection as binary token-level START/END/INSIDE classification rather than span-based prediction, (2) incorporating NP chunking and dependency features that help discontinuous and NOUN-type MWEs identification, and (3) applying oversampling that addresses severe class imbalance in the training data. We confirm the generalization of our method on the STREUSLE dataset, achieving 78.9% F1. These results demonstrate that carefully designed smaller models can substantially outperform LLMs on structured NLP tasks, with important implications for resource-constrained deployments.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Multi-word Expression Identification on CoAM (test)](<https://www.sota2.com/research/sota/multi-word-expression-identification-on-coam-test.md>): Multi-word Expression Identification; CoAM (test); Precision: 70.

- [Multi-word Expression Identification on STREUSLE (test)](<https://www.sota2.com/research/sota/multi-word-expression-identification-on-streusle-test.md>): Multi-word Expression Identification; STREUSLE (test); Precision: 75.5.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2601.19360>)

[PDF](<https://arxiv.org/pdf/2601.19360.pdf>)
