choiminji-020102/ko-comments-emotion-analyzer 🔢📝 → ❓

▶️ 27 runs 📅 Aug 2025 ⚙️ Cog 0.16.5
keyword-extraction korean sentiment-analysis

About

Analyze Korean comments for sentiment classification into negative, positive, and neutral categories (with neutral designated for confidence scores below 0.7), and extract corresponding sentiment-specific keywords.

Example Output

Output

{"keywords":{"neutral":["배고파요"],"negative":["비싸네요"],"positive":["맛있어요"]},"individual_results":[{"id":1,"result":"NEUTRAL"},{"id":2,"result":"POSITIVE"},{"id":3,"result":"NEGATIVE"}]}

Performance Metrics

0.30s Prediction Time
107.70s Total Time
All Input Parameters
{
  "ids": [
    1,
    2,
    3
  ],
  "texts": [
    "배고파요",
    "정말 맛있어요! 추천합니다",
    "가격이 좀 비싸네요"
  ]
}
Input Parameters
ids Type: array
각 텍스트 ID 리스트
texts (required) Type: array
분석 텍스트 리스트
Output Schema

Output

Type: object

Example Execution Logs
입력받은 텍스트 리스트 길이: 3
텍스트 리스트: ['배고파요', '정말 맛있어요! 추천합니다', '가격이 좀 비싸네요']
text_id: 1
text: 배고파요
result: NEUTRAL
text_id: 2
text: 정말 맛있어요! 추천합니다
result: POSITIVE
text_id: 3
text: 가격이 좀 비싸네요
result: NEGATIVE
Version Details
Version ID
55f49ebbdc50dd9565906847fb96d0621a19e1122536df254d995d61437b0d1f
Version Created
September 1, 2025
Run on Replicate →