diff --git a/docs/UserGuide/5-Operation Manual/2-DML Data Manipulation Language.md b/docs/UserGuide/5-Operation Manual/2-DML Data Manipulation Language.md
index 50ea02d0b6c3d..5f153540bf47a 100644
--- a/docs/UserGuide/5-Operation Manual/2-DML Data Manipulation Language.md
+++ b/docs/UserGuide/5-Operation Manual/2-DML Data Manipulation Language.md
@@ -244,36 +244,44 @@ The path after SELECT in GROUP BY statement must be aggregate function, otherwis
-### Last timestamp Query
-In scenarios when IoT devices updates data in a fast manner, users are more interested in the most recent record of IoT devices.
-The LAST query is to return the most recent value of the given timeseries in a time-value pair format.
+### Last point Query
-The SQL statement is:
+In scenarios when IoT devices updates data in a fast manner, users are more interested in the most recent point of IoT devices.
+
+The Last point query is to return the most recent data point of the given timeseries in a three column format.
+
+The SQL statement is defined as:
```
select last [COMMA ]* from < PrefixPath > [COMMA < PrefixPath >]*
```
-which means:
-Query and return the data with the largest timestamp of timeseries prefixPath.path.
+which means: Query and return the last data points of timeseries prefixPath.path.
+
+The result will be returned in a three column table format.
-In the following example, we queries the latest record of timeseries root.ln.wf01.wt01.status:
```
-select last status from root.ln.wf01.wt01 disable align
+| Time | Path | Value |
```
-The result will be returned in a three column table format.
+
+Example 1: get the last point of root.ln.wf01.wt01.speed:
+
```
+> select last speed from root.ln.wf01.wt01
+
| Time | Path | Value |
| --- | ----------------------- | ----- |
-| 5 | root.ln.wf01.wt01.status| 100 |
+| 5 | root.ln.wf01.wt01.speed | 100 |
```
-If the path root.ln.wf01.wt01 has multiple columns, for example id, status and temperature, the following case will return records of all the three measurements with the largest timestamp.
+
+Example 2: get the last speed, status and temperature points of root.ln.wf01.wt01
+
```
-select last id, status, temperature from root.ln.wf01.wt01 disable align
+> select last speed, status, temperature from root.ln.wf01.wt01
| Time | Path | Value |
| --- | ---------------------------- | ----- |
-| 5 | root.ln.wf01.wt01.id | 10 |
+| 5 | root.ln.wf01.wt01.speed | 100 |
| 7 | root.ln.wf01.wt01.status | true |
| 9 | root.ln.wf01.wt01.temperature| 35.7 |
```
diff --git a/docs/zh/UserGuide/5-Operation Manual/2-DML Data Manipulation Language.md b/docs/zh/UserGuide/5-Operation Manual/2-DML Data Manipulation Language.md
index dc4e97e7a8e14..d359ebd1e1fc4 100644
--- a/docs/zh/UserGuide/5-Operation Manual/2-DML Data Manipulation Language.md
+++ b/docs/zh/UserGuide/5-Operation Manual/2-DML Data Manipulation Language.md
@@ -273,35 +273,40 @@ GROUP BY的SELECT子句里的查询路径必须是聚合函数，否则系统将
-### 最近时间戳数据查询
+### 最新数据查询
-对应的SQL语句是：
+SQL语法：
```
select last [COMMA ]* from < PrefixPath > [COMMA < PrefixPath >]*
```
-其含义是：
-查询时间序列prefixPath.path中最近时间戳的数据
+其含义是：查询时间序列prefixPath.path中最近时间戳的数据
+
+结果集为三列的结构
-下面的例子中查询时间序列root.ln.wf01.wt01.status最近时间戳的数据:
```
-select last status from root.ln.wf01.wt01 disable align
+| Time | Path | Value |
```
-结果集为以下的形式返回：
+
+示例 1：查询 root.ln.wf01.wt01.speed 的最新数据点
+
```
+> select last speed from root.ln.wf01.wt01
+
| Time | Path | Value |
| --- | ----------------------- | ----- |
-| 5 | root.ln.wf01.wt01.status| 100 |
+| 5 | root.ln.wf01.wt01.speed | 100 |
```
-假设root.ln.wf01.wt01中包含多列数据，如id, status, temperature，下面的例子将会把这几列数据在最近时间戳的记录同时返回：
+示例 2：查询 root.ln.wf01.wt01 下 speed，status，temperature 的最新数据点
+
```
-select last id, status, temperature from root.ln.wf01 disable align
+> select last speed, status, temperature from root.ln.wf01
| Time | Path | Value |
| --- | ---------------------------- | ----- |
-| 5 | root.ln.wf01.wt01.id | 10 |
+| 5 | root.ln.wf01.wt01.speed | 100 |
| 7 | root.ln.wf01.wt01.status | true |
| 9 | root.ln.wf01.wt01.temperature| 35.7 |
```