A framework for employee appraisals based on inductive logic programming and data mining methods

Aqel, DMA

Abstract

Employee performance appraisal systems are widely regarded as fundamental for evaluating employees’ performance and enhancing organisations’ success. Yet, there is evidence that employees doubt their benefits and fairness, organisations find them difficult to implement and their value is questioned. Although commercial systems that support appraisals have been developed, their focus remains on recording and tracking information, thereby not providing the kind of meaningful and deeper support for appraisals and the goal setting process such as ensuring that the objectives are SMART (specific, measurable, achievable, realistic, time-related) and providing feedback on these objectives.
Developing a supportive employee appraisal system for setting objectives represents a major challenge for computer science since the objectives are unstructured, assessing objectives is a subjective process, and there is no known system for writing effective objectives, providing feedback and supporting the decision making process.
Thus, helping employees to write SMART objectives requires finding the rules of writing these objectives. As the objective sentences are expressed in natural language, natural language processing (NLP) techniques and machine learning may have the potential for supporting the process of setting objectives by first analysing the objectives text and then identifying the rules for writing SMART objectives. More specifically, machine learning methods such as inductive logic programming (ILP), which investigates the inductive acquisition of first-order theories from background knowledge and examples, could be applied to automatically learn the rules.
The process of setting objectives also requires assessing whether the objectives can be met given the available resources and time. Data mining techniques may have the potential to be used for assessing the objectives.
This thesis develops a new framework for employee appraisals. The developed framework supports the process of setting SMART objectives and providing feedback. The framework utilises ILP to learn the grammar rules for writing SMART objectives and applies data mining techniques for assessing the objectives. The framework has been implemented using the ILP system ALEPH as well as prediction and classification rule induction algorithms in the WEKA data mining software. A novel system has been developed based on the proposed framework to show the feasibility of the framework. An empirical evaluation of the developed system has been conducted using a corpus of 300 objective examples and achieved promising results with an overall accuracy of 83%. The thesis also includes the limitations of the developed framework and proposes the potential for further research.

Citation

Aqel, D. A framework for employee appraisals based on inductive logic programming and data mining methods. (Thesis). University of Salford

• Thesis Type: Thesis
• Deposit Date: Mar 7, 2014
• Publicly Available Date: Apr 5, 2016