This paper’s title is Distinguishing Japanese Non-standard Usages from Standard Ones (Aoki et al., EMNLP 2017).
The following is what I prepared for for paper seminar of NLP labs.
I think this paper is interesting to use out-embedding, in here they used out-embedding predicting context word on Skip-gram method.
Tip:
word embedding, in-embedding, out-embedding, distance weight(decay weight)
Note:
This paper focus on non-standard usages of common words on social media. In the context of social media. words sometimes have other usages that ar totally different from their original. so this paper attempt to distinguish non-standard usages on social meadia from standard ones in an unsupervised manner. The their idea is based on mesuaring consistency between the expected meaning of the target word and the given context. For this purpose, They used context embedding derived from word embedding on Skip gram
Download URL:
The paper: Distinguishing Japanese Non-standard Usages from Standard Ones (Aoki et al., EMNLP 2017)
The paper: Distinguishing Japanese Non-standard Usages from Standard Ones (Aoki et al., EMNLP 2017)
Reference
- Paper
- How to use html for alert