Dynamic Pooling and Unfolding Recursive Autoencoders for Paraphrase Detection

By Richard Socher, Eric Hsin-Chun Huang, Jeffrey Pennington, Andrew Y. Ng, Christopher D. Manning. 2011

971 citations recorded by the source corpus.

How researchers see this paper

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

In Natural Language Processing

Citations in the corpus

Recurrent Neural Network Grammars (2016, background): “…neural architectures have been explored in a number of applications, including discriminative parsing (Socher et al., 2013a; Kiperwasser and Goldberg, 2016), sentiment analysis (Tai et al., 2015; Socher et al., 2013b), and sentence representation (Socher et al., 2011; Bowman et al., 2006).”
Long Short-Term Memory-Networks for Machine Reading (2016, neutral): “Models 5-classes 2-classes RAE (Socher et al., 2011) 43.”
Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation (2014, background): “Earlier, a similar encoder–decoder model us-ing two recursive neural networks was proposed in (Socher et al., 2011), but their model was restricted to a monolingual setting, i.e. the model reconstructs an input sentence.”
An efficient framework for learning sentence representations (2018, background): “Socher et al. (2011) proposed recursive autoencoders which encode an input sentence using a recursive encoder and a decoder reconstructs the hidden states of the encoder.”
Skip-Thought Vectors (2015, methodology): “…SDT-RNN [23] 0.7900 0.7304 0.3848 LSTM [22] 0.8528 0.7911 0.2831 Bidirectional LSTM [22] 0.8567 0.7966 0.2736 Dependency Tree-LSTM [22] feats [24] 73.2 RAE+DP [24] 72.6 RAE+feats [24] 74.2 RAE+DP+feats [24] 76.8 83.6 FHS [25] 75.0 82.7 PE [26] 76.1 82.7 WDDP [27] 75.6 83.0 MTMETRICS [28]…”
Distributed Representations of Sentences and Documents (2014, methodology): “The results also show that more advanced methods (such as Recursive Neural Network (Socher et al., 2013b)), which require parsing and take into account the compositionality, perform much better.”
Efficient Estimation of Word Representations in Vector Space (2013, methodology): “ank correlation over the previous best result [31]. The neural network based word vectors were previously applied to many other NLP tasks, for example sentiment analysis [12] and paraphrase detection [28]. It can be expected that these applications can beneﬁt from the model architectures described i”
A Deep Relevance Matching Model for Ad-hoc Retrieval (2016, background): “Although various deep matching models have been proposed under such a general matching problem formalization, most of them have only been demonstrated to be eﬀective on a set of NLP tasks such as paraphrase identiﬁcation and QA [11, 26].”
Grounded Compositional Semantics for Finding and Describing Images with Sentences (2014, methodology): “RNN models (Pollack, 1990; Goller and K¨uchler, 1996; Socher et al., 2011b; Socher et al., 2011a) provided a novel way of combining word vectors for longer phrases that moved beyond simple averaging.”
Convolutional Networks on Graphs for Learning Molecular Fingerprints (2015, methodology): “A recursive neural network [23, 24] is then run from the leaves to the root to produce a fixed-size representation.”
What you can cram into a single $&!#* vector: Probing sentence embeddings for linguistic properties (2018, methodology): “We also train with an AutoEncoder objec-tive (Socher et al., 2011) on Europarl source English sentences.”
Recurrent Continuous Translation Models (2013, background): “…sentiment, paraphrase or dialogue labels, signiﬁcantly beyond the word level and to accurately predict labels for a highly diverse range of unseen phrases and sentences (Grefenstette et al., 2011; Socher et al., 2011; Socher et al., 2012; Hermann and Blunsom, 2013; Kalchbrenner and Blunsom, 2013).”
Ask Me Anything: Dynamic Memory Networks for Natural Language Processing (2015, methodology): “For instance, recursive neural networks have been used for parsing [11], sentiment analysis [2], paraphrase detection [11] and question answering [12] and logical inference [13], among other tasks.”
Towards Universal Paraphrastic Sentence Embeddings (2015, methodology): “…have developed and shared word embeddings trained on large datasets (Collobert et al., 2011; Mikolov et al., 2013; Pennington et al., 2014), and these have been used effectively for many downstream tasks (Turian et al., 2010; Socher et al., 2011; Kim, 2014; Bansal et al., 2014; Tai et al., 2015).”
Reasoning about Entailment with Neural Attention (2015, background): “Despite the success of neural networks for paraphrase detection (e.g. Socher et al., 2011; Hu et al., 2014; Yin and Sch¨utze, 2015), end-to-end differentiable neural architectures failed to get close to acceptable performance for RTE due to the lack of large high-quality datasets.”
SummaRuNNer: A Recurrent Neural Network Based Sequence Model for Extractive Summarization of Documents (2016, methodology): “2014) employed the recursive autoencoder (Socher et al. 2011) to summarize documents, producing best performance on the Opinosis dataset (Ganesan, Zhai, and Han 2010).”
Interaction Networks for Learning about Objects, Relations and Physics (2016, background): “The “graph neural network” [23, 19] is a framework that shares learning across nodes and edges, the “recursive autoencoder” [25] adapts its processing architecture to exploit an input parse tree, the “neural programmer-interpreter” [22] is a composable network that mimics the execution trace of a…”