Lazy-k: Decoding for Constrained Token Classification
This work addresses the challenge of enhancing model accuracy in information extraction tasks, particularly for resource-constrained settings, though it appears incremental as it builds on existing constrained decoding methods.
The paper tackles the problem of improving probabilistic models in structured prediction by combining them with constrained decoding approaches for token classification in information extraction, resulting in significant performance improvements, especially with smaller models, and introducing Lazy-k for flexible decoding time-accuracy trade-offs.
We explore the possibility of improving probabilistic models in structured prediction. Specifically, we combine the models with constrained decoding approaches in the context of token classification for information extraction. The decoding methods search for constraint-satisfying label-assignments while maximizing the total probability. To do this, we evaluate several existing approaches, as well as propose a novel decoding method called Lazy-$k$. Our findings demonstrate that constrained decoding approaches can significantly improve the models' performances, especially when using smaller models. The Lazy-$k$ approach allows for more flexibility between decoding time and accuracy. The code for using Lazy-$k$ decoding can be found here: https://github.com/ArthurDevNL/lazyk.