Extraction of Salient Sentences from Labelled Documents

By Misha Denil, Alban Demiraj, Nando de Freitas. 2014

145 citations recorded by the source corpus.

We present a hierarchical convolutional document model with an architecture designed to support introspection of the document structure. Using this model, we show how to use visualisation techniques from the computer vision literature to identify and extract topic-relevant sentences. We also introduce a new scalable evaluation technique for automatic sentence extraction systems that avoids the need for time consuming human annotation of validation data.

Subfields: Artificial Intelligence

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Explaining Recurrent Neural Network Predictions in Sentiment Analysis (2017, background): “, 2016), or to general deep neural networks for text (Denil et al., 2014; Li et al., 2016a; Arras et al., 2016a; Li et al., 2016b; Murdoch and Szlam, 2017).”
Ask the GRU: Multi-task Learning for Deep Text Recommendations (2016, methodology): “We employ a simple, easy-to-implement tool for analyzing RNN predictions, based on Denil et al. [55] and Li et al. [56].”
Describing Multimedia Content Using Attention-Based Encoder-Decoder Networks (2015, methodology): “able to techniques designed to inspect the representations captured by the model [12], [30], [31].”
Explaining Predictions of Non-Linear Classifiers in NLP (2016, methodology): “Most work dedicated to understanding neural network classiﬁers for NLP tasks (Denil et al., 2014; Li et al., 2015) use gradient-based approaches.”
Dissecting Recall of Factual Associations in Auto-Regressive Language Models (2023, methodology): “We distinguish between sensitivity and saliency (Ancona et al., 2019; Bastings and Filippova, 2020): methods such as Gradient-L2 (Li et al., 2016) show to which inputs the model is sensitive , i.e., where a small change in the input would make a large change in the output, but a method such as Gradi”
Large-Scale Cross-Category Analysis of Consumer Review Content on Sales Conversion Leveraging Deep Learning (2019, unknown): “We implement a method created by Denil et al. (2014) that adapts visualization techniques from computer vision to automatically extract relevant sentences from labeled text data.”
From Group to Individual Labels Using Deep Features (2015, methodology): “The model of Denil et al. [7] is a multi-level convolutional network, which represents a document using a two stage process.”
Explaining and Interpreting LSTMs (2019, methodology): “Besides LRP we also consider gradient-based explanation [21,45,68,13], occlusion-based relevance [46,79], and Contextual Decomposition (CD) [56], as alternative methods.”
Evaluating Explanation Methods for Neural Machine Translation (2020, background): “This baseline metric extends the idea of Arras et al. (2016); Denil et al. (2014) from classiﬁcation tasks to structured prediction tasks like machine translation which are highly dependent on context rather than just key-words.”
Text summarization using unsupervised deep learning (2017, methodology): “In supervised models, Denil et al. (2014) proposed a model based on a convolutional neural network (CNN) to extract candidate sentences to be included into the summary.”