K-NRM (a kernel based neural model for document ranking) 模型的纯numpy实现,并将K-NRM 模型作为特征提取器 结合XGBboost_classifier(分类器) 得到一个优化版本的文本相似度度模型
- Python=3.6
- xgboost=0.90
TODO
- add module to deal train feature
- add xgboost classifier module
- add scritps to run whole model
- finish simple demo