# Selective Encoding for Abstractive Sentence Summarization

## TL;DR

This paper introduces a selective encoding model that enhances sequence-to-sequence frameworks for abstractive sentence summarization, leading to improved performance on multiple datasets.

## Contribution

The paper presents a novel selective gate network that constructs a second-level sentence representation, improving summarization quality over existing models.

## Key findings

- Outperforms state-of-the-art baseline models
- Effective on multiple datasets (Gigaword, DUC 2004, MSR)
- Demonstrates the benefit of selective encoding in summarization

## Abstract

We propose a selective encoding model to extend the sequence-to-sequence framework for abstractive sentence summarization. It consists of a sentence encoder, a selective gate network, and an attention equipped decoder. The sentence encoder and decoder are built with recurrent neural networks. The selective gate network constructs a second level sentence representation by controlling the information flow from encoder to decoder. The second level representation is tailored for sentence summarization task, which leads to better performance. We evaluate our model on the English Gigaword, DUC 2004 and MSR abstractive sentence summarization datasets. The experimental results show that the proposed selective encoding model outperforms the state-of-the-art baseline models.