Pseudo Siamese Network for Few-shot Intent Generation

TL;DR

This paper introduces a Pseudo Siamese Network that generates labeled data for few-shot intent detection, improving performance by modeling sentence components with transformer-based VAEs.

Contribution

The novel Pseudo Siamese Network architecture effectively generates labeled data for few-shot intent detection, addressing annotation scarcity.

Findings

Achieves state-of-the-art results on real-world datasets

Effectively models action and object components in sentences

Improves few-shot intent detection performance

Abstract

Few-shot intent detection is a challenging task due to the scare annotation problem. In this paper, we propose a Pseudo Siamese Network (PSN) to generate labeled data for few-shot intents and alleviate this problem. PSN consists of two identical subnetworks with the same structure but different weights: an action network and an object network. Each subnetwork is a transformer-based variational autoencoder that tries to model the latent distribution of different components in the sentence. The action network is learned to understand action tokens and the object network focuses on object-related expressions. It provides an interpretable framework for generating an utterance with an action and an object existing in a given intent. Experiments on two real-world datasets show that PSN achieves state-of-the-art performance for the generalized few shot intent detection task.