Unsupervised method for measuring smart home service quality through gap analysis and dependency parsing

Abstract

Smart home is one of emerging technology issue in recent time. With increasing interest in smart home environment, companies subjectively developed technology related to smart home service, but because the destination of development was incompatible with customer perception and expectation, smart home adoption was not favorable as much as they expected. Hence, provider of smart home service needs to check how consumers accept their smart home service from the consumer's point of view. Therefore, this paper purposes to evaluate service quality comparing between consumer expectation and perception of service using gap analysis by employing dependency parsing, unsupervised sentiment analysis method, with case study of providing smart home service using reviews evaluating one of the front runner companies of smart home industry in South Korea, A. In this research, we found advantages in extracting more specific service factors which make customer dissatisfaction or satisfaction, of using dependency parsing instead of using traditional sentiment analysis method. The result of dependency parsing analysis can suggest the priorities to improve service factors and confirm demand of customer recognized from customers' perspective. Moreover, we check about verification that dependency parsing is a proper method to analysis customers' review data.