Learning to Compose Task-Specific Tree Structures

2017-07-10 · Jihun Choi, Kang Min Yoo, Sang-goo Lee

For years, recursive neural networks (RvNNs) have been shown to be suitable for representing text into fixed-length vectors and achieved good performance on several natural language processing tasks. However, the main drawback of RvNNs is that they require structured input, which makes data preparation and model implementation hard. In this paper, we propose Gumbel Tree-LSTM, a novel tree-structured long short-term memory architecture that learns how to compose task-specific tree structures only from plain text data efficiently. Our model uses Straight-Through Gumbel-Softmax estimator to decide the parent node among candidates dynamically and to calculate gradients of the discrete decision. We evaluate the proposed model on natural language inference and sentiment analysis, and show that our model outperforms or is at least comparable to previous models. We also find that our model converges significantly faster than other models.

📄 PDF Abstract BibTeX arXiv:1707.02786

Code (1)

jihunchoi/unsupervised-treelstm pytorch

Tasks

Natural Language Inference Sentiment Analysis

Learning to Compose Dynamic Tree Structures for Visual Contexts

2018-12-05 · CVPR 2019 6 · Kaihua Tang, Hanwang Zhang, Baoyuan Wu, Wenhan Luo 외

We propose to compose dynamic tree structures that place the objects in an image into a visual context, helping visual reasoning tasks such as scene graph generation and visual Q&A. Our visual context tree model, dubbed …

Graph Generation Panoptic Scene Graph Generation Reinforcement Learning Scene Graph Generation+2

Learning to Compose Words into Sentences with Reinforcement Learning

2016-11-28 · Dani Yogatama, Phil Blunsom, Chris Dyer, Edward Grefenstette 외

We use reinforcement learning to learn tree-structured neural networks for computing representations of natural language sentences. In contrast with prior work on tree-structured models in which the trees are either prov…

reinforcement-learning Reinforcement Learning Reinforcement Learning (RL)

Learning to Compose over Tree Structures via POS Tags

2018-08-18 · Gehui Shen, Zhi-Hong Deng, Ting Huang, Xi Chen

Recursive Neural Network (RecNN), a type of models which compose words or phrases recursively over syntactic tree structures, has been proven to have superior ability to obtain sentence representation for a variety of NL…

POS Semantic Composition Sentence Sentence Classification+1

Parsing with Traces: An $O(n^4)$ Algorithm and a Structural Representation

2017-07-13 · Jonathan K. Kummerfeld, Dan Klein

General treebank analyses are graph structured, but parsers are typically restricted to tree structures for efficiency and modeling reasons. We propose a new representation and algorithm for a class of graph structures t…

Constituency Parsing Missing Elements

Parsing with Traces: An O(n4) Algorithm and a Structural Representation

2017-01-01 · TACL 2017 1 · Jonathan K. Kummerfeld, Dan Klein

Question Answering