Recursive Neural Networks with Bottlenecks Diagnose (Non-)Compositionality
This work addresses the challenge of measuring compositionality in NLP for researchers, though it is incremental as it builds on existing recursive neural models.
The authors tackled the problem of quantifying compositionality in natural language data by introducing a bottleneck compositionality metric (BCM) using recursive neural networks with bottlenecks, and demonstrated its effectiveness in distinguishing compositional from non-compositional samples for tasks like arithmetic expressions and sentiment classification.
A recent line of work in NLP focuses on the (dis)ability of models to generalise compositionally for artificial languages. However, when considering natural language tasks, the data involved is not strictly, or locally, compositional. Quantifying the compositionality of data is a challenging task, which has been investigated primarily for short utterances. We use recursive neural models (Tree-LSTMs) with bottlenecks that limit the transfer of information between nodes. We illustrate that comparing data's representations in models with and without the bottleneck can be used to produce a compositionality metric. The procedure is applied to the evaluation of arithmetic expressions using synthetic data, and sentiment classification using natural language data. We demonstrate that compression through a bottleneck impacts non-compositional examples disproportionately and then use the bottleneck compositionality metric (BCM) to distinguish compositional from non-compositional samples, yielding a compositionality ranking over a dataset.