How to Encode Domain Information in Relation Classification
Authors:
Elisa Bassignana,
Viggo Unmack Gascou,
Frida Nøhr Laustsen,
Gustav Kristensen,
Marie Haahr Petersen,
Rob van der Goot,
Barbara Plank
Abstract:
Current language models require a lot of training data to obtain high performance. For Relation Classification (RC), many datasets are domain-specific, so combining datasets to obtain better performance is non-trivial. We explore a multi-domain training setup for RC, and attempt to improve performance by encoding domain information. Our proposed models improve > 2 Macro-F1 against the baseline set…
▽ More
Current language models require a lot of training data to obtain high performance. For Relation Classification (RC), many datasets are domain-specific, so combining datasets to obtain better performance is non-trivial. We explore a multi-domain training setup for RC, and attempt to improve performance by encoding domain information. Our proposed models improve > 2 Macro-F1 against the baseline setup, and our analysis reveals that not all the labels benefit the same: The classes which occupy a similar space across domains (i.e., their interpretation is close across them, for example "physical") benefit the least, while domain-dependent relations (e.g., "part-of'') improve the most when encoding domain information.
△ Less
Submitted 21 April, 2024;
originally announced April 2024.
