从零开始撸论文-英文润色篇（工具版）

英文润色

一篇好的论文，文字关是无法避免的。
对于自己动手修改论文的人来说，一些好的润色工具对于克服文字关可以起到事半功倍的效果，让自己的文字变得更加地道或者说更加准确。
关于这一块，我分享一下自己常用的论文润色工具，以及使用流程。

准备工具：

首先看一下常用工具，
wordviceAI：英文润色工具（检查错误用的，会提出语境的建议，以后可能收费，替代项有languageToo，有道翻译的英文批改）
谷歌翻译：中译英
deepin：英译中（感觉英译中比谷歌好）

流程

样例段：
From the viewpoint of multi-view learning, MESS is a multi-view regression method with serial updating and safe labeling.
In each iteration, the same confident instance in all views tends to be included the training set of next view.
This updating method can better guarantee the quality of selected instances than single view.
Afterwards, the safe label will be learned from other views before being included into training set.
This finely suits the manner of multi-view that exploring the unlabeled data by redundant views.
第一步：wordviceAI检查错误与初步润色

修改后：
第二步：英转中，调整中文表意

修改后：

第三步：中转英


针对中译英联合修改：

润色检查定冠词等错误：

最后版本：
From the viewpoint of multi-view learning, MESS is a multi-view regression method with serial updates and safety labels.
In each training, the regressor will tend to pick one or more high-confidence instances.
After that, MESS will learn the safety label of confidence instances with the help of other views.
Confidence instances and safety labels will be included in the next view training set.
This update method guarantees the quality of the selected instance better than a single view.
It is also finely suitable for the idea of multi-view methods, that is, to explore unlabeled data through redundant views.