diff --git a/doc/en/mobilizer.md b/doc/en/mobilizer.md
index a3297fc..c54867f 100644
--- a/doc/en/mobilizer.md
+++ b/doc/en/mobilizer.md
@@ -55,3 +55,26 @@ So which page will display by default? Feed or web page?
 - Show feed view if you enable `Download web page when reading` in `feed settings` and network is NOT available.
 
 When the article content uses for some cases, like `Share content` or play it, it is also following these rules. 
+
+# FeedMe Mobilizer Selector (supported since v4.1)
+`FeedMe Mobilizer` uses a simple algorithm to parse the body content of a URL address. Generally, more accurate analysis can be obtained when the text content is relatively large. However, due to the variety of web pages, always happens this simple algorithm parsing incorrectly.
+
+In CSS, selectors are used to select certain element(s) of HTML. Starting from v4.1, FeedMe will simply use some selectors to help `FeedMe Mobilizer` locate content parts more accurately. I call this new feature `FeedMe Mobilizer Selector`.
+
+## advantage
+- More accurate analysis of the text content of the article
+- Less parsing time
+
+## shortcoming
+- HTML/CSS knowledge required
+- Need to know how to locate a node in HTML through the browser
+
+## quick guide
+1. Choose one feed you need to set the `FeedMe Mobilizer selector`. It usually does not provide full text and `FeedMe Mobilizer` cannot parse it correctly.
+2. Open any article from this feed in a desktop browser.
+3. Find the smallest node containing the text by looking at the element node, and find the ID or class that can be used for positioning.
+4. Open FeedMe, open the subscription management screen of this feed, select `FeedMe` in `Mobilizer`, and the `FeedMe Mobilizer Selector` section will be displayed.
+5. Enter the ID or class found for positioning. There are two situations:
+    - What you found is ID, for example `content_body`, enter `#content_body`
+    - What you found is class, for example `content_body`, enter `.content_body`
+6. At this point, the settings have been completed. You can open the web again to see if the parsing is correct. If the web has been opened, it needs to be reloaded and parsed through the `Reload page` of the article interface.
diff --git a/doc/zh/mobilizer.md b/doc/zh/mobilizer.md
index d59c05b..4220885 100644
--- a/doc/zh/mobilizer.md
+++ b/doc/zh/mobilizer.md
@@ -54,3 +54,26 @@ Web的内容通过源链接重新抓取，并通过Mobilizer解析。一般情
 - 显示 feed 页面，如果你在`订阅源设置`对话框启用了`阅读时下载 web 页面`但网络不可用。
 
 当文章内容在某些功能中使用时，比如`全文分享`或者播放，它同样会遵循以上规则。 
+
+# FeedMe Mobilizer 选择器 （从v4.1开始支持）
+`FeedMe Mobilizer` 使用了一个简单的算法来解析一个URL地址中的正文内容。一般在正文内容比较多时可以获得比较准确的解析。但由于网页各式各样，总有这个简单算法解析错误的时候。
+
+在CSS中，选择器用来选择HTML的某一些元素。从v4.1开始，FeedMe会简单的使用一些类选择器的思路来帮助`FeedMe Mobilizer`更准确的定位内容部分。我把这个新功能叫做`FeedMe Mobilizer 选择器`。
+
+## 优点
+- 更加准确的解析文章正文内容
+- 更少的解析时间
+
+## 缺点
+- 需要HTML/CSS知识
+- 需要知道如何通过浏览器来定位HTML中某个节点
+
+## 快速指南
+1. 确定你需要设置`FeedMe Mobilizer 选择器`的订阅源，它一般是没有提供全文输出并且`FeedMe Mobilizer`不能正确解析。
+2. 在桌面浏览器打开这个订阅源的任意一篇文章。
+3. 通过查看元素节点，找到包含正文的最小节点，找到可以用来定位的ID或者class
+4. 打开FeedMe，打开这个订阅源的订阅管理界面，`Mobilizer`选择`FeedMe`，这时就会显示`FeedMe Mobilizer Selector`部分
+5. 输入找到的用来定位的ID或者class，这里分2种情况：
+    - 找到的是ID，比如 `content_body`，输入 `#content_body`
+    - 找到的是class，比如 `content_body`，输入 `.content_body`
+6. 到此为止，设置已经完成。你可以再次打开web看解析是否正确。如果web已经打开过，需要通过文章界面的`Reload page`来重新加载并解析。