Recent statistics showed that 75% of applicants nowadays are filtered out from applicant pools for a position by Applicant Tracking System, due to the inflexibility of machines in understanding differemt resume formats and synonyms. Lander is a text-mining based tool that helps the student increase their chance of passing the initial resume screening round by optimizing their resume using a keyword-suggesting system. Moreover, Lander will also match a candidate to a compatible job by matching the experiences in their resume to a job description, and in return, increase the chance of getting accepted into such job.
With the detrimental effect of Covid-19, the job market has become extremely competitive for the last 3 years, especially for newly graduated candidates. During the summer of my freshman year, I was working as an intern for Talent Acquisition Service Office of a renowned bank back in Vietnam. In there, I have witnessed so many fresh candidates who apply to a myriad of jobs without actually knowing what they want to do or what they are good at. Then, when I came back to America, a lot of seniors back then also tell me that they had the same problem. At that time, I realize that the struggle in job hunting is a worldwide problem.
Thus, my dream since then was to create a platform where fresh candidates can find what is the most suitable job they should apply to based on their past experiences and interests. In addition, I want to help them tailor their resume based on the job description of their dream job to enhance their chance of getting through the initial scanning round.
Python will be the main language that I use to implement this artifact, and I will deliver the result via streamlit or netlify depends on which one has a better interface for this tools.
In order to match job description with resume, I will use keywords extraction feature from gensim, phrase.matcher from spacy, and cosine-similarity from sklearn. Before that, I will try my best to group word/skills that are coherently matched/equal/refer to each other by training a model on job description using gpt-2 or tensorflow.
In order to recommend which key word the student should incorporate to their resume, I extract the keywords in the job description using gensim and I will rank the importance of the keyword by an algorithm that I will design later on
[1] Babu, Deepak. "Leveraging Technology to Improve Candidate Experience." NHRD Network Journal 11.2 (2018): 29-31.
[2] SINGH, Ms SUPRIYA, and SP SINGH. "ADVANTAGES & DISADVANTAGES OF E-RECRUITMENT." Advance Management Practices in Business: 43.
[3] Sippy, Mishika, et al. "ResumeScan: Application Tracking and Career Prediction Model." (2021).