tcet-opensource / resume-screener Goto Github PK

• A resume scoring algortihm
• Resume strength factor

Name the file as extract_user_domain.py

• The User resume must be scrapped and stored in JSON format
• Use Pypdf2json to scrape the user profile

Perform the above tasks and store the file in utils/

Name the file github_analyser.py

• Import the user profile data from utils/github_scrapper.py
• Do pre-processing on the github user data and transform it
• get the job profile description
• perform content-based analysis using cosine similarity on it
• output matching score and grade

Keep this model inside models/

Describe the improvement

• creating a API route recommend_courses.py inside the routes folder to return json data of the course_recommendation.py model output

• the route is tested

Why do we need this?

to integrate models to the resume screener frontend

Do you have a proposed solution?

yes creating routes lol what else launch a nuclear weapon ?

Do you want to work on this feature?

• I am willing to implement this feature.

Terms and Conditions

• I agree to follow this project's Code of Conduct

• create show_user_details route
• create a show_user_data.html
• display basic details of user like name, email, number, education/college etc

• Make a file './utils/ats_functions.py' which has class 'Ats_Functions' and make the function under this class
  • Check for the presence of image in the resume (resume should not be present in a good resume)
  • Return if resume is acceptable or not (personal image is present or not)

• user dashboard completed
• user dashboard button generate option done

• created ML model in models/ for skill_recommendation
• created ML model in models/ for course_recommendation

• goto routes/recommend_skills.py and complete the route
• pass user data of skills and GET skill recommendations
• return recommend_skills.json

• Make a file './utils/ats_functions.py' which has class 'Ats_Functions' and make the function under this class
  • Use computer vision to check the percentage of all colors present in resume
  • Accept only a certain range of colors.
    • Slightly darker shades of white
    • Slightly lighter shades of black
    • Other colors only upto a certain threshold
  • Return if the resume is acceptable or not (percengates of colors present)

• show user suggestions function & route
• created a rule based ML model in models/ that will detect and analsye user resume to std ATS format

• pdf2json conversion module done

• Make a file './utils/ats_functions.py' which has class 'Ats_Functions' and make the function under this class
  • Check for number of characters/words in resume
  • Ensure that the number does not exceed a specific threshold
  • Return weather resume is acceptable (character count)

Name the file identify_user_domain.py

• Import the user profile data from utils/extract_resume_json.py
• Do pre-processing on the user data and transform it
• get the job relevant skills description
• perform rule based mapping on skills and profession
• output matching profession

Keep this model inside models/

• Make the changes in 'utils' folder under the name 'resume_data_to_json_tree
  • Create class and a function within the python file
  • Clean the extracted data from OCR
  • Format the data in a proper json tree structure
  • Return the json file in './data' folder

• Make a file './utils/ats_functions.py' which has class 'Ats_Functions' and make the function under this class
  • Analyze the document to for alignment and spacing using parsers (or OCR parsers)
  • Score the document based on the alignment (no. lines out or place/total lines or any other metric)

• create user sub-section scores for each section acc to ATS std format
• create an overall scoring function to rate the user resume out of 100 score or 100%
• create scores.html and link to index.html

• recommend course acc to user recommended skills
• add slider to select number of recommended courses

• goto routes/recommend_courses.py and complete the route

• pass user data of skills/domain and GET course recommendations

• return recommend_courses.json

• Change theme
• Add background (Tcet open source theme)
• Add Tcet open source logo
• add animations or any effects if necessary

• Create a app.py for the flask app

• initialize a dashboard
• create an outline in st components

• create recommend_skills function & route it
• recommend_courses function & route it
• render recommendations.html and link to index.html page

we will be showing all these things in user dashboard in index.html. you can use models/ to import the machine learning models into ur flask routes.

Describe the improvement

• creating a API route recommend_skills.py inside the routes folder to return json data of the skill_recommendation.py model output

• the route is tested

Why do we need this?

to integrate models to the resume screener frontend

Do you have a proposed solution?

yes creating routes lol what else launch a nuclear weapon ?

Do you want to work on this feature?

• I am willing to implement this feature.

Terms and Conditions

• I agree to follow this project's Code of Conduct

• create a structure for the project
• create routes/
• create models/
• create data/
• create templates/ & static/
• other files for the flask app

boiler plate for the project should be ready

Name the file as linkedln_scrapper.py

• User will input their profile URL
• The User profile must be scrapped and stored in JSON format
• Use API to scrape the user profile
• Use a Web crawler to scrape the profile page details

Perform the above tasks and store the linkedln_scrapper.py in utils/

Name the file linkedln_analyser.py

• Import the user profile data from utils/linekdln_scrapper.py
• Do pre-processing on the linkedln user data and transform it
• get the job profile description
• perform content-based analysis using cosine similarity on it
• output matching score and grade

Keep this model inside models/

• upload resume route done
• render upload.html done
• Get resume data route done

The upload_resume route is already present add functionality similarly the get_resume_data route is also present add functionality to it. Create a upload.html to take user input [POST] and then scrape the data from resume using resume_parser module of pyresparser u can see the reference branch for reference how it was built.

Name the file as github_scrapper.py

• User will input their profile URL
• The User profile must be scrapped and stored in JSON format
• Use API to scrape the user profile
• Use a Web crawler to scrape the profile page details

Perform the above tasks and store the github_scrapper.py in utils/

Name the file as model_resume_format.py

• The Resume uploaded by the user should be converted into the specific format (json) which makes the extraction of data easy
• Build model using spacy or any other nlp tool which can format the unstructured data into structured.
  Perform the above tasks and store the file in models/

ResumeParser is a module in pyresparser that specifically built to parse and extract pdf data into text components.

• ResumeParser function
• Test & output some basic user data

Describe the improvement

Utilise the custom OCR tool to extract the resume contents.

Why do we need this?

To integrate our OCR tool into the project.

Do you have a proposed solution?

Simple extraction and structuring of the resume content

Do you want to work on this feature?

• I am willing to implement this feature.

Terms and Conditions

• I agree to follow this project's Code of Conduct

A function that allows upload and display of resume

• File upload upto 10-50 MB
• Resume display
• Save uploaded resume locally

• Make a file './utils/ats_functions.py' which has class 'Ats_Functions' and make the function under this class
  • Check for bullet points in resume
  • Make sure no paragraphs present in the resume
  • Return weather resume is acceptable (paragraphs present or not)

• Make the changes in 'model' folder under the name 'resume_extraction_OCR.py'
  • Create class and a function within the python file to extract data
  • Use tesseract (or any suitable method) to extract the characters, sections and raw text data
  • Return extracted data in a dictionary with format {file_name: "RESUME_FILE_NAME", data:["EXTRACTED DATA IN A LIST"]}

• Display user level
• Display user skills/tech
• Identify the User domain
• Generate & display recommendations

Name the file as resume_ats_scan.py

• The Resume uploaded by the user can be converted to Image
• To find space percentage and then categorize it into Excellent, Good, Need Improvement

Perform the above tasks and store the file in models/

• modify existing #8 and create route
• improve the existing solution by creating a scoring factor based on skills and experience
• create a route in routes -> resume_scorer.py that takes user data and processes a score

• convert the user data into sub json

eg : resume contains experience, skills, achievements, internships ,etc divide them into sub parts and store in json format for easier convenience while performing actions on those specific data. Import these later while creating diff routes & functions

Describe the improvement

We need to create a custom OCR tool for our purposes. Training an OCR model

Why do we need this?

Since pre-existing tools are not cutting it for our application, we decided on developing a custom one.

Do you have a proposed solution?

Training an OCR model using computer vision

Do you want to work on this feature?

• I am willing to implement this feature.

Terms and Conditions

• I agree to follow this project's Code of Conduct

• Make a file './utils/ats_functions.py' which has class 'Ats_Functions' and make the function under this class
  • Check count of personal pronoun usage in resume corpus
  • Reject resume if personal pronouns above a certain threshold/present at all
  • Return weather resume is acceptable

I noticed that this repo doesn't have contribution guidelines file.

• initialize app
• create a basic streamlit interface

• from the current user skills & interests find the user domain create route
• show user domain in dashboard
• identify user level of experience by scoring user skills & internships

• save file in 'models' folder

• Use ats functions to calculate ATS score
• Calculate formatting score using computer vision
• return the aggregate of both the score as final score