Prediction of the apply rate of the postings based on the job characteristics

Job portals and job listings such as Glassdoor, Indeed, and LinkedIn use different data mining techniques and machine learning algorithms to provide the best job recommendations based on a candidate’s preferences. Job recommendations are not only based on the preferences set by candidates. Rather, there are other parameters that need to be considered as well, such as skills required by the job, relocation being provided or not, searched keywords, visa sponsorship, etc. The job recommendations provided by these job recommendation systems play a significant role in the company’s growth and the “Apply Rate” or “Click Through Rate” (CTR) for a particular job posting. CTR is a metric used to measure the success of an online advertisement campaign and how it impacts advertisement rank and quality score. In this research, the apply rate is considered as the measure of whether a user will apply to that particular job or posting based on different characteristics such as job title, job description, popularity of job, job listing matches, location, time period of job listing, etc. The results of this research can be analysed in order to improve future job searches, click rates, and job matches in order to attract more users in an online advertisement system. Amongst the data mining techniques used in this work for prediction of CTR (apply rate), the random forest classifier accuracy was 94.39%, which was higher than the accuracy values of 93.97% for the decision tree classifier and 93.15% for the K-Nearest Neighbour (KNN) classifier