diff --git a/.github/workflows/ai-review.yml b/.github/workflows/ai-review.yml
new file mode 100644
index 0000000..dcb8c38
--- /dev/null
+++ b/.github/workflows/ai-review.yml
@@ -0,0 +1,38 @@
+# This workflow enables developers to call PR-Agents `/[actions]` in PR's comments and upon PR creation.
+# Learn more at https://www.codium.ai/pr-agent/
+# This is v0.2 of this workflow file
+
+name: PR-Agent
+
+on:
+  pull_request:
+    types: [opened]
+  issue_comment:
+    types: [created]
+  workflow_dispatch:
+
+permissions:
+  issues: write
+  pull-requests: write
+
+jobs:
+  pr_agent_job:
+    runs-on: ubuntu-latest
+    name: Run pr agent on every pull request
+    if: ${{ github.event.sender.type != 'Bot' }}
+    steps:
+      - name: PR Agent action step
+        id: pragent
+        uses: Codium-ai/pr-agent@main
+        env:
+          OPENAI_KEY: ${{ secrets.OPENAI_KEY }}
+          # OPENAI_ORG: ${{ secrets.OPENAI_ORG }} # optional
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          # PINECONE.API_KEY: ${{ secrets.PINECONE_API_KEY }}
+          # PINECONE.ENVIRONMENT: ${{ secrets.PINECONE_ENVIRONMENT }}
+          GITHUB_ACTION_CONFIG.AUTO_DESCRIBE: true
+          GITHUB_ACTION_CONFIG.AUTO_REVIEW: true
+          GITHUB_ACTION_CONFIG.AUTO_IMPROVE: true
+          PR_DESCRIPTION.EXTRA_INSTRUCTIONS:      'Please use Japanese.'
+          PR_REVIEWER.EXTRA_INSTRUCTIONS:         'Please use Japanese.'
+          PR_CODE_SUGGESTIONS.EXTRA_INSTRUCTIONS: 'Please use Japanese.'
diff --git a/simirality/__pycache__/main.cpython-311.pyc b/simirality/__pycache__/main.cpython-311.pyc
index 5e54373..4c83cfd 100644
Binary files a/simirality/__pycache__/main.cpython-311.pyc and b/simirality/__pycache__/main.cpython-311.pyc differ
diff --git a/simirality/main.py b/simirality/main.py
index 22d5cd7..b0165cc 100644
--- a/simirality/main.py
+++ b/simirality/main.py
@@ -29,34 +29,71 @@ class Words(BaseModel):
     assignmentWord: str
     words: List[str]
 
-
 @app.post("/similarity")
 async def similarity(reqWords: Words):
     assignmentWord = reqWords.assignmentWord
     words = reqWords.words
 
+    try:
+        highscore, highscore_word = calcuSimilarity(assignmentWord, words)
+    except:
+        # 例外処理
+        print("Error: 類似度の計算に失敗しました。課題が存在しない可能性があります。")
+        return {"similarity": 0}
+
+    return {"similarity": highscore, "highscoreWord": highscore_word}
+
+# 類似度の計算
+def calcuSimilarity(assignmentWord, words):
     assignmentWord_synset = wn.synset(f"{assignmentWord}.n.01")
 
     highscore = 0
+    highscore_word = ""
+
     for word in words:
         print(f"word: {word}")
         try:
-            word_synset = wn.synset(f"{word}.n.01")
-            similarity = assignmentWord_synset.wup_similarity(word_synset)
-            print(f"{word}: {similarity}")
+            similarity = calculate(assignmentWord_synset, word, assignmentWord)
 
-            if assignmentWord == word:
-                similarity = 1.0
-
-            if similarity is None:
-                print(f"'{assignmentWord}' と '{word}' の類似度を計算できません。")
+            print(f"similarity: {similarity}")
 
             if similarity > highscore:
                 highscore = similarity
+                highscore_word = word
         except:
             print(f"Error: '{word}' はWordNetに存在しません。")
 
-    return {"similarity": highscore}
+        # WordNet にない形式を検索
+        wordByMorphy = wn.morphy(word)
+        print(f"wordByMorphy: {wordByMorphy}")
+        if wordByMorphy is not None:
+            try:
+                similarityByMorphy = calculate(assignmentWord_synset, wordByMorphy, assignmentWord)
+
+                print(f"similarity: {similarityByMorphy}")
+
+                if highscore < similarityByMorphy:
+                    highscore = similarityByMorphy
+                    highscore_word = word
+
+            except:
+                print(f"Error: '{word}' はWordNetに存在しません。")
+
+    return highscore, highscore_word
+
+def calculate(assignmentWord_synset, word, assignmentWord):
+    if assignmentWord == word:
+        return 1.0
+
+    word_synset = wn.synset(f"{word}.n.01")
+
+    similarity = assignmentWord_synset.wup_similarity(word_synset)
+
+    if similarity is None:
+        print(f"'{assignmentWord}' と '{word}' の類似度を計算できません。")
+        return 0
+
+    return similarity
 
 
 # MEMO こんな感じでPOSTリクエストを送る