Team Members 

Piyush Sharan - 234009304   

Manisha Panda - 334003453 

Janvita Reddy - 634005984  

Neha Manghnani - 234009317 

Links

Demo Video:  https://www.youtube.com/watch?v=tLy7v4JJtQI

Github: https://github.com/pi-sharan/JobHunter

JobHunter aims to build a Job Recommendation System that takes a multi-faceted approach towards understanding a user’s past work experience and combining it with the jobs they have applied for in the recent past. The goal is to create a centralized platform catering to both job seekers and providers, offering a seamless experience akin to swiping left or right to indicate preference for recommended job opportunities.This tool can be utilized by job seekers to find jobs relevant to their experience and skill set. In the current job market, having a tool that filters through a number of job advertisements and recommends job positions based on skills and experience would add value to the job seeker by helping them spend less time finding relevant jobs and focusing on building their profile for their desired jobs. JobHunter would help the job seeker in finding the next right job to add to their profile and help them in building their career trajectory. 

Motivation

Job search tools exist such as LinkedIn Jobs and Indeed. However, they are more of a job search engine, rather than being a recommendation system. These tools have job advertisements posted and the user has to search for the positions they want to apply for. Moreover, the user has to read the job description to identify if the job is right for them before applying. This demands a lot of time and effort from the job seeker. Our tool, JobHunter, would be a recommendation engine where we would recommend relevant jobs to job seekers based on their previous experience and current skill set. We would also recommend jobs that users of similar experience and skill sets have applied to. This gives them more reach to jobs relevant to their area of expertise

Challenges

For the user’s (job seekers) history, only the job titles are provided, we would have to fetch the job description and skills relevant to the job title. Some of the job titles have multiple meanings i.e. Software Engineer could mean both Backend, Frontend, or Fullstack developer but they are explicitly mentioned in the dataset. Additionally, the scope of the job is not defined (skill sets/tools used for the job are missing). Similarly, we noticed that some users have a history of unrelated jobs. Lastly, there is a cold start problem where some users have not provided their previous employment history by choice or they do not have one, in this case, we would have to look at alternatives to recommend jobs

Approach

Results