It Research Proposal – How Data Mining & Decision Tree Can Improve Personnel Selection and Human Capital

...

Using an input variable and a respective output variable we can construct a decision tree. There a several algorithms such CART (Classification and Regression Tree), CHAID (chi-square Automatic Interaction Detector). We will use CHAID as the data mining tool to explore the relationship among the input employee profile and target variables of behaviour of work like turnover, job assessment etc.

Approach

Constructing a framework for HR for data mining to find out the relationship between personnel profiles and behaviour of the job. With this approach, we can find hidden information from datasets and be useful for decision making of finding the right talent. The framework for data mining should have the following steps:

- Problems and objective: The first step of building the framework, we need to identify the correct problem and addressing the objective

- Collecting Data: Collecting the right data from various sources and enriching the data into analysable formats.

- Evaluating model for Data Mining: When the model is ready it needs to be evaluated before we can use for decision support systems.

- Analysing and evaluating the model: All the results should be interpreted and needs to be assessed accordingly. Useful patterns and trends can be summarized into the decision support systems.

- Extracting and Interpreting: The results should be interpreted and justified.

- Using the extracted data: The interpreted data can be the basis for decision support system to generate the desired strategies we need to hire the right talent.

Conclusion

The only way a company can maintain its competitive advantage is by human capital. This research can help develop a framework for data mining to extract important data for identifying relationships between employee profile and their behaviour at work. If a company can have effective personnel selection process they can grab the most suitable talent at initial stages to improve employee turnover and maximise performance. Even further the data that we mined can be used for useful strategies such as career development, redesigning job, rotation of job. This is possible by knowing the demographics of an employee such as age, sex, qualification and work experience. To further go ahead with this research, we can collect more input variables such as address and grades of school/university. We have used decision tree for data mining in this research as its easier and accurate. We can also use alternate techniques in data mining such as neutral network that can studied, compared and implemented for interrelationships among input personal variables and behaviour of work. This can further be added in systems like HRIS (Human Resource Information System) as initial screening for applicants that will reduce work of recruiters and save costs.

References

Chien, C., & Chen, L. (2008, January). How Data Mining and Decision Tree can improve personnel selection and improve human capital. Retrieved from http://www.sciencedirect.com/science/article/pii/S0957417406002776

...