DialogConv: A Lightweight Fully Convolutional Network for Multi-view Response Selection

TL;DR

DialogConv introduces a lightweight, fully convolutional network for multi-view response selection in dialogue systems, achieving comparable accuracy to larger models with significantly reduced size and faster inference on CPU and GPU.

Contribution

The paper presents a novel convolutional architecture, DialogConv, that models dialogues in multiple views for efficient and effective response selection, outperforming existing models in size and speed.

Findings

DialogConv is 8.5x smaller than state-of-the-art models.

It is 79.39x faster on CPU and 10.64x faster on GPU.

Maintains competitive response selection accuracy.

Abstract

Current end-to-end retrieval-based dialogue systems are mainly based on Recurrent Neural Networks or Transformers with attention mechanisms. Although promising results have been achieved, these models often suffer from slow inference or huge number of parameters. In this paper, we propose a novel lightweight fully convolutional architecture, called DialogConv, for response selection. DialogConv is exclusively built on top of convolution to extract matching features of context and response. Dialogues are modeled in 3D views, where DialogConv performs convolution operations on embedding view, word view and utterance view to capture richer semantic information from multiple contextual views. On the four benchmark datasets, compared with state-of-the-art baselines, DialogConv is on average about 8.5x smaller in size, and 79.39x and 10.64x faster on CPU and GPU devices, respectively. At the same time, DialogConv achieves the competitive effectiveness of response selection.