A curated list of jobs keeping in mind your cognitive, sensory and mobile impairments, powered by data and delivered to your fingertips. Unleash your true potential, don't let the world hold you back.
- Lack of clarity for job compatibility for employees with impairments
- Difficulty in finding job listings for the less tech savvy
An NLP powered algorithm that uses past employement data to scout and provide you with the most suitable job listings for you based on your impairments, delivered to you through the Indeed job portal or via email
- LV - Low Vision
- HI - Hearing Impaired
- MPD - Mild Physical Disability
- PD - Physical Disability
- CP - Cerebral Palsy
indeedscraper.py
- initial web scraper to populate job databasejob_smartFilter.py
- prepares our NLP pipeline, and standardizes the job posting input to an analysable formatruleset.py
- generates a rule binding between jobs and the disability classes based on past employment data found through web url scrapingworld_cluster_predictor.py
- creates the NLP model, saves it, and reads through the job posting database, preparing it for impairment tag additionmodel_ODE.py
- loads our NLP model, runs it on the test data, adds the tag for the impairment suitability and exports it to a spreadsheet form.StochasticPredictor.ipynb
- Predicts job suitability likelihood using a Stochastic Gradient Descent Regression algorithm over a Bag of Words model of hand picked data set- Along with these scripts, some additional scripts were written to help clean and regulate our databases, but they have not been added here as they are standard pandas dataframe cleaning functions.
- The .txt files contain standardised data generated from past employment data to help us find key words for jobs secured by people with different impairments
300_job_listings.xls
- 300 job postings from Indeed, cleaned upRated_650JobPostings_cleaned.xlsx
- 650 job postings from Indeed, cleaned upindeed_ODE.csv
- output from our Gensin/FastText NLP algorithm, with job suitability rating for each impairment tagindeed_ODE_average.csv
- output that combines the results of both our Stochastic Gradient Descent Regression algorithm along with the Gensin/FastText NLP algorithm, with job suitability rating for each impairment tag