Developing Interpretive Text Analysis for Corporate Culture Measurement

Corporate culture is crucial to an organization's success, as it influences teamwork, productivity, efficiency, and customer service ability. However, there has yet to be a consensus on measuring corporate culture. Most tools used for measuring corporate culture are self-reported questionnaires, which are prone to faking. Projective evaluation methods, however, are less popular due to time-intensive administration and scoring, and the need for uniformity in interpretation. These disadvantages of projective assessment could be reduced through machine-based text analysis. We developed a text-analysis projective assessment for corporate culture using a case study of one Indonesia's State-Owned Enterprise, which has recently undergone significant organizational changes, including digital transformation. The processes of developing the tools are operationalizing the variables, creating a corpus for training the model, and evaluating the model. We found that the combination of Label Powerset and Linear SVC performs best among all combinations of multi-label classification approach and algorithm. Moreover, we discuss the possibilities of future works to improve the tool.