This repository applies BERT for named entity recognition on resumes. The goal is to find useful information present in resume.
pip3 install -r requirements.txtTo train model use:
python3 train.pyoptional arguments:
-e epochs number of epochs
-o path output path to save model state
python3 app.pyAPI will be live at localhost:5000 endpoint predict.
curl --location --request POST 'http://localhost:5000/predict' --form 'resume=@/resume-path.pdf'Output:
{
"entities": [
{
"entity": "Name",
"start": 3,
"end": 19,
"text": "Ayush Srivastava"
},
{
"entity": "Designation",
"start": 22,
"end": 35,
"text": "Web Developer"
},
{
"entity": "Degree",
"start": 50,
"end": 56,
"text": "B.Tech"
},
{
"entity": "Years of Experience",
"start": 72,
"end": 73,
"text": "3"
},
{
"entity": "College Name",
"start": 937,
"end": 943,
"text": "JSSATE"
},
{
"entity": "Degree",
"start": 955,
"end": 961,
"text": "B.Tech"
},
{
"entity": "Graduation Year",
"start": 964,
"end": 968,
"text": "2016"
},
{
"entity": "Skills",
"start": 1188,
"end": 1219,
"text": "○ Designing (UI/UX & Photoshop)"
},
{
"entity": "Skills",
"start": 1221,
"end": 1251,
"text": "○ Web Development (HTML & CSS)"
},
{
"entity": "Skills",
"start": 1287,
"end": 1323,
"text": "ReactJS, Gatsby, jQuery, JavaScript,"
},
{
"entity": "Skills",
"start": 1326,
"end": 1360,
"text": "HTML, CSS, Materialize, Bootstrap,"
},
{
"entity": "Skills",
"start": 1373,
"end": 1378,
"text": "Flask"
},
{
"entity": "Skills",
"start": 1402,
"end": 1416,
"text": "D3, Matplotlib"
},
{
"entity": "Skills",
"start": 1433,
"end": 1454,
"text": "Google Cloud Platform"
},
{
"entity": "Skills",
"start": 1481,
"end": 1488,
"text": "C++, C,"
},
{
"entity": "Skills",
"start": 1491,
"end": 1497,
"text": "Python"
},
{
"entity": "Skills",
"start": 1511,
"end": 1521,
"text": "Oracle SQL"
}
]
}Links: Actual Resume Full Response
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent