diff --git a/.gitignore b/.gitignore
index 2bffc7a..84e2a23 100644
--- a/.gitignore
+++ b/.gitignore
@@ -5,3 +5,4 @@ Gemfile.lock
 doc/*
 pkg/*
 coverage/*
+vendor/*
diff --git a/CHANGELOG.md b/CHANGELOG.md
new file mode 100644
index 0000000..96c2cc1
--- /dev/null
+++ b/CHANGELOG.md
@@ -0,0 +1,31 @@
+# Changelog
+
+## v0.2.0
+
+### Added
+
+- Add `tokenizer` option to `Document` class
+
+  The value is an object with a `tokenize` method that accepts a string and returns an array of `Token` instances. 
+
+  For example, to use [natto](https://rubygems.org/gems/natto) instead of [unicode_utils](https://rubygems.org/gems/unicode_utils) for Japanese, install MeCab (`brew install mecab`), and then:
+
+  ```ruby
+  require 'natto'
+
+  class Tokenizer
+    def initialize
+      @nm = Natto::MeCab.new
+    end
+
+    def tokenize(text)
+      @nm.enum_parse(text).map do |node|
+        Token.new(node)
+      end
+    end
+  end
+
+  document = TfIdfSimilarity::Document.new("こんにちは世界", tokenizer: tokenizer)
+  ```
+
+- Add `to_s` method to `Token` class, to use less memory than chaining `lowercase_filter` with `classic_filter`