Recursive Deep Models for Semantic Compositionality Over a Sentiment Treebank
Paper’s reference in the IEEE style?
R. Socher, A. Perelygin, J. Y. Wu, J. Chuang, C. D. Manning, A. Y. Ng, and C. P. Potts, “Recursive Deep Models for Semantic Compositionality Over a Sentiment Treebank,” in EMNLP, 2013.
How did you find the paper?
If applicable, write a list of the search terms you used.
Was the paper peer reviewed? Explain how you found out.
Does the author(s) work in a university or a government-funded research institute? If so, which university or research institute? If not, where do they work?
The researchers all work at Stanford university
What does this tell you about their expertise? Are they an expert in the topic area?
All are experienced researchers in the field of NLP and artificial intelligence
What was the paper about?
This paper discusses the development of recursive neural tensor networks (RNTN) and the Stanford sentiment treebank for sentiment analysis. The treebank was built by Stanford using Amazon's mechanical turk to tag sentences from the Rotten Tomatoes movie review data set and it can be used online here.
The production of this extensive network was then used to train the RNTN.
Again, simplistically, a recursive tensor network is a kind of neural network which operates recursively over a structure. For NLP the structure is the word tree. RNTNs are RNNs which use a consistent tensor composition for all nodes of the tree2)https://en.wikipedia.org/wiki/Recursive_neural_network.
RNTNs 'learn' the semantic structure of language and perform well in sentiment analysis, and in accounting for word negation and other linguisitcly difficult issues.
The following tree comes from entering "I don't like maths, I love it" in the tree bank site.
If applicable, is this paper similar to other papers you have read for this assignment? If so, which papers and why?
If applicable, is this paper different to other papers you have read for this assignment? If so, which papers and why?
What do these similarities and differences suggest? What are your observations? Do you have any new ideas? Do you have any conclusions?
This is the original paper introducing the Stanford tree bank and RNTNs' RNTNs are an advanced artificial neural network for use in sentiment analysis and are able to predict sentiments with an accuracy in excess of 80% as well and interpret linguistic modifiers reasonably well.
Over the last 2-3 years there are ongoing advancements in the use of machine learning to improve NLP performance.
It is likely there is still a significant level of improvement possible.
This question is to be answered after your critical analysis is completed. Which sections (if any) of your critical analysis was this paper cited in?
References [ + ]