Status : Verified
| Personal Name | Mangubat, Arra Mae M. |
|---|---|
| Resource Title | Enhancing online recruitment: a targetted solutioning through multilabel classification |
| Date Issued | 20 July 2024 |
| Abstract | This capstone project aimed to develop a machine learning model using data from 2022 to 2023 to help in predicting the errors in the application submitted by an aspiring financial advisor. The existing controls employed by the company were also assessed to recommend feasible solutions to improve the current recruitment process. A high error rate in the applications leads to a longer average turnaround time to complete the process which prevents a company from growing its manpower. This study used R software’s MLR package to create a multilabel classification task given that there can be multiple error classifications that can apply to a single data point. The model was trained until its performance metrics reached a level acceptable to the business. The results showed that 40% of the errors in the two-year period are encountered in a single requirement, while the remaining 60% is comprised by 18 other error classifications. While there are existing controls and guidelines to inform the applicants about the recruitment process and requirements, these are not enough given the consistently high error rate. The study concludes with a set of recommended action items involving user education and platform enhancements that the company may implement both in the short- and long-run. The data-based results and recommendations offer valuable insights to the company in order to improve the online recruitment process. Lastly, the machine learning model used in this study can be re-applied to other multilabel classification tasks, both inside and outside the company. |
| Degree Course | Master of Technology Management |
| Language | English |
| Keyword | Multilabel classification; Online recruitment |
| Material Type | Thesis/Dissertation |
Preliminary Pages
282.66 Kb
Category : C - Confidential information of a third-party is embedded.
Access Permission : Limited Access