go-dict-tagging

Annotate natural sentences based on a dictionary.

Example

1. request

http://127.0.0.1:8080/tag?statement=我有糖尿病，在吃格华止，可以同时吃奥利司他来减肥吗？

2. response

{
    "code": 1,
    "msg": "",
    "result": {
        "奥利司他": {
            "keyword": "奥利司他",
            "positions": [
                {
                    "start": 17,
                    "end": 20
                }
            ],
            "dictwords": {
                "medicine": [
                    {
                        "dict": "medicine",
                        "word": "奥利司他胶囊",
                        "index": [
                            "奥利司他",
                            "奥利司他胶囊"
                        ],
                        "data": {
                            "category": "西药",
                            "categoryId": 0,
                            "isControl": true,
                            "isHighRisk": true,
                            "medicineType": 0
                        }
                    },
                    {
                        "dict": "medicine",
                        "word": "奥利司他片",
                        "index": [
                            "奥利司他",
                            "奥利司他片"
                        ],
                        "data": {
                            "category": "西药",
                            "categoryId": 0,
                            "isControl": true,
                            "isHighRisk": true,
                            "medicineType": 0
                        }
                    }
                ]
            }
        },
        "格华止": {
            "keyword": "格华止",
            "positions": [
                {
                    "start": 8,
                    "end": 10
                }
            ],
            "dictwords": {
                "brand": [
                    {
                        "dict": "brand",
                        "word": "格华止",
                        "index": [
                            "格华止"
                        ],
                        "data": {
                            "category": "西药"
                        }
                    }
                ]
            }
        },
        "糖尿": {
            "keyword": "糖尿",
            "positions": [
                {
                    "start": 2,
                    "end": 3
                }
            ],
            "dictwords": {
                "symptom": [
                    {
                        "dict": "symptom",
                        "word": "糖尿",
                        "index": [
                            "糖尿"
                        ],
                        "data": {
                            "symptomId": 0
                        }
                    }
                ]
            }
        },
        "糖尿病": {
            "keyword": "糖尿病",
            "positions": [
                {
                    "start": 2,
                    "end": 4
                }
            ],
            "dictwords": {
                "disease": [
                    {
                        "dict": "disease",
                        "word": "糖尿病",
                        "index": [
                            "糖尿病"
                        ],
                        "data": {
                            "diseaseId": 0
                        }
                    }
                ]
            }
        }
    },
    "micros": 672
}

API

uri	method	content-type	parameter	usage
/	GET			HTML API Document
/tag	GET/POST	application/json	statement	Annotate for statement
/put	POST	multipart/form-data	file	Upload single {dictname}.json file
/reload	GET/POST			Rebuild the index tree based on the dictionary file
/info	GET/POST			Display the dictionary information that the current index tree depends on

Usage

Windows

1. Download the Windows program archive from the release page.
2. Extract the archive.
3. (Optional)Edit your own dictionary entries in the data directory.
4. Double-click to run dict_tagging.exe.
5. Access in the browser: http://localhost:8080/tag?statement=your_statement

Linux

1. Download the linux program archive from the release page.
2. Extract the archive.
3. (Optional)Edit your own dictionary entries in the data directory.
4. ./dict_tagging
5. Access in the browser: http://your-server-ip:8080/tag?statement=your_statement

Docker

cd /path/to/dict_tagging_dir
docker build -t dict_tagging:1.0.2 .
docker run -d -p 8080:8080 dict_tagging:1.0.2