Dynamic Feature Generation Network for Answer Selection

TL;DR

This paper introduces DFGN, a novel network that dynamically generates and filters sentence-level features using attention mechanisms, significantly improving answer selection performance.

Contribution

The paper presents a new dynamic feature generation network that enhances sentence representations with attention-based features and automatic filtering, advancing answer selection methods.

Findings

Outperforms state-of-the-art baselines on multiple datasets

Demonstrates strong retrieval and interpretative capabilities

Provides detailed analysis of feature importance

Abstract

Extracting appropriate features to represent a corpus is an important task for textual mining. Previous attention based work usually enhance feature at the lexical level, which lacks the exploration of feature augmentation at the sentence level. In this paper, we exploit a Dynamic Feature Generation Network (DFGN) to solve this problem. Specifically, DFGN generates features based on a variety of attention mechanisms and attaches features to sentence representation. Then a thresholder is designed to filter the mined features automatically. DFGN extracts the most significant characteristics from datasets to keep its practicability and robustness. Experimental results on multiple well-known answer selection datasets show that our proposed approach significantly outperforms state-of-the-art baselines. We give a detailed analysis of the experiments to illustrate why DFGN provides excellent retrieval and interpretative ability.