- repair --> 增强，丰富，
The kkk-nearest neighbor algorithm enhances the feature representation of the target sample by obtaining neighbors.
Most of them take account of label correlation which plays an importance role in LDL.
–>
Most of them take account of label correlation, which plays an important role in LDL.
- 归纳总结
有几类（全局和局部）
标签相关性根据范围可以分为全局和局部相关性。
全局相关性利用流行学习[]、[]。。。
局部相关性利用…[],[]
还有一些工作[]同时考虑了二者。
标签相关性根据关联程度可以分为标签之间的关联性和标签和样本之间的关联性。
)
)
)
)