magicpieh28 / Paper-Summary

1 stars 0 forks source link

Double Embeddings and CNN-based Sequence Labeling for Aspect Extraction(2018) #28

Open magicpieh28 opened 5 years ago

magicpieh28 commented 5 years ago

about this paper

Author: Hu Xu, Bing Liu, Lei Shu, Philip S. Yu Link: https://www.aclweb.org/anthology/P18-2094.pdf

model

일반 엠베딩 concat 도메인 특화 엠베딩 -> 복수의 CNN 레이어 -> FC -> Softmax

experiments

parameters

results

conclusion

aspect 용어가 A 그리고 B와 같은 형식일 때 A만 추출되고 B는 추출되지 않는다. 이는 다음 과제로 남겨둔다.

code

https://github.com/howardhsu/DE-CNN

next

Think globally, embed locally-locally linear meta-embedding of words. 2017. IJCAI-18 Unsupervised cross-domain word representation learning. 2015. IJCNLP Convolutional sequence to sequence learning. 2017. Facebook Deep multi-task learning for aspect term extraction with memory interaction. 2017. EMNLP Coupled multi-layer attentions for co-extraction of aspect and opinion terms. 2017. AAAI Lifelong domain word embedding via meta-learning. 2018. AAAI