Detecting Fake Job Postings Using NLP

Within the job market, plenty of juniors fall prey to fake job postings which can lead them to be scammed and lose interest in the field, or in the worst case, lose hope in their dreams or even get caught up in criminal acts. The problem regarding this is prevalent within the domain. With how ambiguous the data science professional world has been with its job titles, it has also led to a plethora of people, junior and experienced alike, ending up misunderstanding a job posting and getting a role they didn’t have in mind. This project aims to minimize such occurrences.

As individuals navigate the job market, they often face the potential pitfalls of fraudulent job advertisements, which may lead to exploitation or scamming by deceptive entities. There's also the risk of aligning with companies that lack a comprehensive understanding of the roles they're advertising.

This project aims to allow people to glean from the job posting whether it is a fake job or also a true one that shouldn’t be trusted. It also has as a goal to identify the patterns of words perceived within these aforementioned job postings that should be avoided.

• Create a model to correctly identify the words that would represent a high risk of fake job postings.
• Create a model to classify a job posting as true and safe or not.
• Create a word cloud visualization to identify the most common words among fake job postings.

NLP, Data Wrangling, Wordcloud, Deep Learning, Webscraping, Clustering