Papers by Muyao Niu
A Label Informative Wide & Deep Classifier for Patents and Papers (D19-1)
Copied to clipboard
| Challenge: | Existing methods for classification of patents and papers are manual and limited . et al., a chinese research team has developed a label-informative classification model . |
| Approach: | They propose a label-informative classifier based on the Wide & Deep structure . they train on millions of patents and transfer to papers by developing distant-supervised training set and domain-specific features. |
| Outcome: | The proposed model performs comparable to the state-of-the-art model used in industry on patents and papers. |