BYOC: Personalized Few-Shot Classification with Co-Authored Class Descriptions

TL;DR

This paper introduces BYOC, a personalized few-shot text classification method where users co-author class descriptions with an LLM, enabling high-accuracy classifiers with minimal data and user interaction.

Contribution

The paper presents a novel interactive approach for few-shot classification that leverages co-authored class descriptions with an LLM, reducing data requirements and enhancing personalization.

Findings

Achieves 82% of large dataset model performance with only 1% data.

End-users can effectively build personalized classifiers.

Personalized classifiers reach 90% accuracy, outperforming state-of-the-art.

Abstract

Text classification is a well-studied and versatile building block for many NLP applications. Yet, existing approaches require either large annotated corpora to train a model with or, when using large language models as a base, require carefully crafting the prompt as well as using a long context that can fit many examples. As a result, it is not possible for end-users to build classifiers for themselves. To address this issue, we propose a novel approach to few-shot text classification using an LLM. Rather than few-shot examples, the LLM is prompted with descriptions of the salient features of each class. These descriptions are coauthored by the user and the LLM interactively: while the user annotates each few-shot example, the LLM asks relevant questions that the user answers. Examples, questions, and answers are summarized to form the classification prompt. Our experiments show that our approach yields high accuracy classifiers, within 82% of the performance of models trained with significantly larger datasets while using only 1% of their training sets. Additionally, in a study with 30 participants, we show that end-users are able to build classifiers to suit their specific needs. The personalized classifiers show an average accuracy of 90%, which is 15% higher than the state-of-the-art approach.