CKIP Service

Web service for ckiplab/ckiptagger

Preparation

• Download model files and put into data folder

Start service

1. Start the service using docker-compose

   docker-compose up -d
   

   If you want to rebuild the image, add --build flag when startup

   docker-compose up --build -d
   

2. Service is now on port 5005

Stop service

1. Stop the service using docker-compose

   docker-compose down
   

Endpoint

• Main
  • method: POST
  • route: /
  • parameter
    • sentence_list: sentence list for CKIP tagging, split multiple sentences by linebreak(\n)

Test CKIP Tagger

1. Send request using curl

   curl -X POST localhost:5005 -F $'sentence_list=土地公有政策?？還是土地婆有政策。.\n最多容納59,000個人,或5.9萬人,再多就不行了.這是環評的結論.'

2. Get the response like the following one

   JSON Response

   {
       "sentences": [
           {
               "segments": [
                   {
                       "word": "土地公有",
                       "pos": "VH"
                   },
                   {
                       "word": "政策",
                       "pos": "Na"
                   },
                   {
                       "word": "?",
                       "pos": "QUESTIONCATEGORY"
                   },
                   {
                       "word": "？",
                       "pos": "QUESTIONCATEGORY"
                   },
                   {
                       "word": "還是",
                       "pos": "Caa"
                   },
                   {
                       "word": "土地",
                       "pos": "Na"
                   },
                   {
                       "word": "婆",
                       "pos": "Na"
                   },
                   {
                       "word": "有",
                       "pos": "V_2"
                   },
                   {
                       "word": "政策",
                       "pos": "Na"
                   },
                   {
                       "word": "。",
                       "pos": "PERIODCATEGORY"
                   },
                   {
                       "word": ".\\n",
                       "pos": "FW"
                   },
                   {
                       "word": "最多",
                       "pos": "Da"
                   },
                   {
                       "word": "容納",
                       "pos": "VJ"
                   },
                   {
                       "word": "59,000",
                       "pos": "Neu"
                   },
                   {
                       "word": "個",
                       "pos": "Nf"
                   },
                   {
                       "word": "人",
                       "pos": "Na"
                   },
                   {
                       "word": ",",
                       "pos": "COMMACATEGORY"
                   },
                   {
                       "word": "或",
                       "pos": "Caa"
                   },
                   {
                       "word": "5.9萬",
                       "pos": "Neu"
                   },
                   {
                       "word": "人",
                       "pos": "Na"
                   },
                   {
                       "word": ",",
                       "pos": "COMMACATEGORY"
                   },
                   {
                       "word": "再",
                       "pos": "D"
                   },
                   {
                       "word": "多",
                       "pos": "D"
                   },
                   {
                       "word": "就",
                       "pos": "D"
                   },
                   {
                       "word": "不行",
                       "pos": "VH"
                   },
                   {
                       "word": "了",
                       "pos": "T"
                   },
                   {
                       "word": ".",
                       "pos": "PERIODCATEGORY"
                   },
                   {
                       "word": "這",
                       "pos": "Nep"
                   },
                   {
                       "word": "是",
                       "pos": "SHI"
                   },
                   {
                       "word": "環評",
                       "pos": "Na"
                   },
                   {
                       "word": "的",
                       "pos": "DE"
                   },
                   {
                       "word": "結論",
                       "pos": "Na"
                   },
                   {
                       "word": ".",
                       "pos": "PERIODCATEGORY"
                   }
               ],
               "entities": [
                   {
                       "word": "59,000",
                       "type": "CARDINAL",
                       "start": 24,
                       "end": 30
                   },
                   {
                       "word": "5.9萬",
                       "type": "CARDINAL",
                       "start": 34,
                       "end": 38
                   }
               ]
           }
       ]
   }

Alternative

If you want to run this service without Docker, you can follow this steps after data folder is ready.
However, we highly recommend using Docker Compose.

1. Install required packages

   pip3 install -r requirements.txt

2. Run

   uvicorn app.main:app --host=0.0.0.0 --port=5005