# Classify or Select: Neural Architectures for Extractive Document Summarization

arXiv: https://arxiv.org/abs/1611.04244 · HTML: https://mlsanity.com/paper/1611.04244/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 14 Nov 2016.

## Built on by (most cited first, 9 of 9)

- [Data Augmentation for Abstractive Query-Focused Multi-Document Summarization](https://mlsanity.com/paper/2103.01863/) (2021)
- [An Extractive-and-Abstractive Framework for Source Code Summarization](https://mlsanity.com/paper/2206.07245/) (2022)
- [Controlled Text Reduction](https://mlsanity.com/paper/2210.13449/) (2022)
- [Fairness-aware Summarization for Justified Decision-Making](https://mlsanity.com/paper/2107.06243/) (2021)
- [A Condense-then-Select Strategy for Text Summarization](https://mlsanity.com/paper/2106.10468/) (2021)
- [Multi-Review Fusion-in-Context](https://mlsanity.com/paper/2403.15351/) (2024)
- [Efficient Extractive Text Summarization for Online News Articles Using Machine Learning](https://mlsanity.com/paper/2509.15614/) (2025)
- [Topic-Aware Encoding for Extractive Summarization](https://mlsanity.com/paper/2112.09572/) (2021)
- [Extractive Summarization for Arabic Documents Using SAraBERT with a Semantic Siamese Similarity Evaluation Metric](https://mlsanity.com/paper/2608.20964/) (2026)

## Abstract (arXiv)

We present two novel and contrasting Recurrent Neural Network (RNN) based architectures for extractive summarization of documents. The Classifier based architecture sequentially accepts or rejects each sentence in the original document order for its membership in the final summary. The Selector architecture, on the other hand, is free to pick one sentence at a time in any arbitrary order to piece together the summary. Our models under both architectures jointly capture the notions of salience and redundancy of sentences. In addition, these models have the advantage of being very interpretable, since they allow visualization of their predictions broken up by abstract features such as information content, salience and redundancy. We show that our models reach or outperform state-of-the-art supervised models on two different corpora. We also recommend the conditions under which one architecture is superior to the other based on experimental evidence.