Dissecting Span Identification Tasks with Performance Prediction

TL;DR

This paper analyzes span identification tasks like NER and chunking by predicting model performance based on task properties, providing insights into how task features influence model effectiveness and guiding architecture choices.

Contribution

It introduces a performance prediction model for span ID tasks, identifying key task properties affecting performance and analyzing how model and task interactions influence results.

Findings

Span frequency significantly impacts LSTM performance.

CRFs improve results when spans are infrequent and boundaries are non-distinctive.

A large-scale experiment supports performance prediction across various span ID tasks.

Abstract

Span identification (in short, span ID) tasks such as chunking, NER, or code-switching detection, ask models to identify and classify relevant spans in a text. Despite being a staple of NLP, and sharing a common structure, there is little insight on how these tasks' properties influence their difficulty, and thus little guidance on what model families work well on span ID tasks, and why. We analyze span ID tasks via performance prediction, estimating how well neural architectures do on different tasks. Our contributions are: (a) we identify key properties of span ID tasks that can inform performance prediction; (b) we carry out a large-scale experiment on English data, building a model to predict performance for unseen span ID tasks that can support architecture choices; (c), we investigate the parameters of the meta model, yielding new insights on how model and task properties interact to affect span ID performance. We find, e.g., that span frequency is especially important for LSTMs, and that CRFs help when spans are infrequent and boundaries non-distinctive.