Papers by Yunli Wang
EuroGames16: Evaluating Change Detection in Online Conversation (L18-1)
Copied to clipboard
| Challenge: | a new method for detecting salient changes from on-line conversations is needed . linguistic preprocessing and time series are used to build a time series . |
| Approach: | They propose a framework for detecting salient changes from on-line conversations . they use linguistic preprocessing to build a time series and change point detection algorithms to detect salient change. |
| Outcome: | The proposed method can detect salient changes in on-line conversations with high accuracy. |
Real-time Change Point Detection using On-line Topic Models (C18-1)
Copied to clipboard
| Challenge: | Existing methods for detecting events from publicly available data streams such as twitter have been used to model topics from large corpora. |
| Approach: | They propose to use on-line Latent Dirichlet Allocation to model topic shifts and on-lines change point detection algorithms to detect when significant changes occur. |
| Outcome: | The proposed algorithm yields F-scores up to 52% on the detection of real-life changes from social media data streams. |
Harnessing Pre-Trained Neural Networks with Rules for Formality Style Transfer (D19-1)
Copied to clipboard
| Challenge: | Existing studies normalize informal sentences with rules, but they introduce noise if we use them in a naive way. |
| Approach: | They propose to harness rules into a state-of-the-art neural network that is typically pretrained on massive corpora. |
| Outcome: | The proposed method can be used to generate a state-of-the-art on a small dataset. |
Formality Style Transfer with Shared Latent Space (2020.coling-main)
Copied to clipboard
| Challenge: | Existing approaches for formality style transfer use neural networks for sentence generation, but the dataset for formal style transfer is considerably smaller than translation corpora. |
| Approach: | They propose a new approach for formality style transfer using shared latent space and two auxiliary losses. |
| Outcome: | The proposed approach outperforms baselines in various settings, especially when limited data is available. |