Recursive Deep Models for Semantic Compositionality Over a Sentiment Treebank

Рекурсивные глубокие модели для семантической композицией по дереву тональности
Christopher D. Manning, Richard Socher, Andrew Y. Ng, Christopher E. Potts, Alex Perelygin, Jean Y. Wu, Jason Chuang
2013-01-01

recursive deep modelrecursive neural networksemantic compositionalitysentiment analysissentiment treebank
Richard Socher, Alex Perelygin, Jean Wu, Jason Chuang, Christopher D. Manning, Andrew Ng, Christopher Potts. Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing. 2013.
1
Created a Sentiment Treebank dataset that provides sentiment labels for phrases within parse trees, enabling fine-grained supervised learning.
2
Demonstrated that recursive neural models trained on the Sentiment Treebank improve sentence- and phrase-level sentiment classification compared to baselines.
3
Introduced recursive deep models for modeling semantic compositionality in sentences using tree-structured representations.
4
Showed that modeling compositional structure via parse trees captures sentiment interactions (e.g., negation) better than bag-of-words approaches.

Recursive deep models (tree-structured neural networks) applied to a sentiment treebank

Modeling semantic compositionality for sentiment analysis—how recursive deep models compose word/phrase meanings to predict sentiment at phrase and sentence levels

Publication Details
Publication Date
2013-01-01
Journal
Publisher
ISSN
Cited by
6863
Access Type
Author Information
Authors
Christopher D. Manning
Richard Socher
Andrew Y. Ng
Christopher E. Potts
Alex Perelygin
Jean Y. Wu
Jason Chuang
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%