Resume parser with ner using state of art in deep learning with transformers specifically roberta.This project is a personal project and not ready for production use but can decently perform parsing on the resumes and extract the keywords and match it with the best possible entitites which i have defined below.
The link for trained model is link to model .I did not added it on github because size was exceeding github size parameter.
There are seven types of entity possible for a given set of words:
- Job title : This entity represents the the type of job which users wants
- skill : This represents the importants skills which users possess
- experience : This represents the job of the user in previous company and it timeline
- org : This represents the set of companies ehich user has worked previously or working now.
- tool : This represents the software tools used by user
- Degree : This represents which degree user has taken for example B.Tech,M.tech,MBA etc
- Educ : This represents the college in which user studied.
I have create a docker container so that it can work on any system because i am tired on people saying not working on my system
- docker should be installed in your machine
- Then download the repo
- Build the docker file using docker build -t resume_docker . (If it does not work use sudo command)
- Then run the docker file using docker run -d -p 5000:5000 resume_docker
- And on this url link you can upload your resume and then the model will give parsed output after sometime so that you will be able to see important keyword in the resume.
Created a complete ci cd pipleine with docker, aws ecs, aws ecr and github actions to automate the workflow of the environment
For more detailed information you can visit my blog at link