Analysis with Deeply Learned Distributed Representations of Variable Length Texts

@inproceedings{Hong2015AnalysisWD,
  title={Analysis with Deeply Learned Distributed Representations of Variable Length Texts},
  author={James Hong and Michael Fang},
  year={2015}
}
Learning good semantic vector representations for phrases, sentences and paragraphs is a challenging and ongoing area of research in natural language processing and understanding. In this project, we survey and implement several deeplearning and deep-learning-inspired approaches and evaluate these algorithms on several sentiment-labeled datasets and analysis tasks. In doing so, we demonstrate new state-of-the-art performance on the IMDB Large Movie Review Dataset [5] using highly-tuned… CONTINUE READING