Employees Opinion Mining – Value of Structured and Unstructured Content Analytics in a Bank

Abstract:

Extensive literature presents the impact of employee satisfaction and the need for organizations to understand the employees voice of the employees, attitude and concerns. In a bank, client facing employees are constantly interacting in a mandatory, complex transactional environment dealing with private information and strict regulations. Information Systems provides mandatory functionality for core banking, customer relationship management systems, specialized systems for transactional and operational tasks. Communication speed, information availability and quality are key in client interaction while employee's messages and attitude against IT complex environment becomes extremely valuable. This paper focuses on evaluating the value of opinions captured as unstructured data, analyzing 586 respondents of survey in open ended questions using text analytics. Extracting key concepts, relationships and performing a quantitative research that evaluates potential analytical model candidates that combines both structured and unstructured data. The exploratory research was based on Technology Acceptance Model; unstructured data processing techniques were used for extracting the factors contributing to Perceived Ease of Use and Perceived Utility constructs. Data was collected for entire IT environment, while open-ended questions focused on best and worst applications, along with reasons for that classifications. Evaluated models included linear models, decision trees and neural nets, providing evaluative comparison criteria. The results suggest similar causality within all methods, while provides additional insights on employee's perception from concept associations.