diff --git a/README.md b/README.md
index 57be645..c6f36d2 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,6 @@
 # v2dat
 
-A cli tool that can unpack v2ray data packages, also known as `geoip.dat` and `geosite.dat`.
+A cli tool that can unpack v2ray data packages (also known as `geoip.dat` and `geosite.dat`) to text files.
 
 ## Usage
 
@@ -10,6 +10,29 @@ v2dat unpack geosite [-d output_dir] [-f tag[@attr]...]... geosite_file
 ```
 
 - If `-d` was omitted, the current working dir `.` will be used.
-- If no filter `-f` was given. All tag will be unpacked.
+- If no filter `-f` was given. All tags will be unpacked.
 - If multiple `@attr` were given. Entries that don't contain any of given attrs will be ignored.
-- Unpacked text files will be named as `<geo_file>_<suffix>.txt`.
\ No newline at end of file
+- Unpacked text files will be named as `<geo_filename>_<filter>.txt`.
+
+## Unpacked IP Data
+
+Unpacked IP text files contain a list of CIDRs.
+
+```text
+2.16.33.76/32
+2.19.128.0/20
+2.20.32.0/22
+```
+
+## Unpacked Domain Data
+
+`geosite` contains four types of domain rule expression: `domain`, `keyword`, `regexp`, `full`. Each expression can have several attributes `@attr`. More info about `geosite` can be found in [here](https://github.com/v2fly/domain-list-community).
+
+`v2dat` will split type and expression with a `:`. But omits the `domain` prefix and attributes.
+
+```text
+google.com
+keyword:google
+regexp:www\.google\.com$
+full:www.google.com
+```
\ No newline at end of file