Publication

Abstract : Online hiring and unpredictable job movement in the IT sector provides a need to analyze the pattern of movement of software professionals. Career trajectory is the trace of the list of jobs shifted by the professional through his career including the key skills acquired. Understanding the career trajectories of employees is a crucial tool for recruiting an employee to provide better stability in the industry. Career trajectory data is abundantly available and needs to be processed for extracting crucial features like educational qualification, work experience, and skill set acquired before each job shift. Career trajectories are forever evolving considering digitalization and enormous amounts of data generated. LinkedIn is being one such social media platform for job search, and it contains a lot of data that could be leveraged to understand the problem of career trajectory analysis. Data for the study has been extracted using web scraping from LinkedIn. Analyzing and recognizing trends and patterns in everlasting data require graphical data visualization. Graphical representation aids in the rapid processing of huge amounts of data and can assist in the making of predictions and well-informed judgments. Community detection algorithm is applied over the dataset to identify the employee communities available and identify relationships between them. This paper analyzes performance of several community detection algorithms on career trajectory graphs using the Girvan–Newman modularity score. Based on the analysis, Louvain algorithm has achieved a better modularity score of 0.85, and Girvan–Newman has the least score of 0.78.