Desenvolvimento de uma extensão de navegador para análise de credibilidade de anúncios de emprego no linkedin utilizando aprendizagem de máquina e processamento de linguagem natural (NLP)

Abstract

The increasing incidence of fraudulent job advertisements on online platforms has become a significant challenge for both candidates seeking new professional opportunities and employers aiming to recruit safely and efficiently. This phenomenon undermines trust in digital recruitment practices, exposing users to financial, emotional, and reputational risks. In response to this scenario, the present study proposes the development of a browser extension designed to assist LinkedIn users in efficiently evaluating the credibility of job advertisements. The adopted methodology comprises documentary research and the application of user surveys to gather requirements and identify user needs. The proposed solution further involves the development of a machine learning model, trained on a labelled dataset, capable of identifying patterns and characteristics indicative of fraudulent job postings. The expected outcomes include strengthening security within the online recruitment environment by providing candidates with a practical tool to support more informed decision-making, thereby helping to mitigate the risks associated with exposure to fraudulent advertisements.