From Naive Expectation to Realistic Progress: Government Applications of Big Data on Public Opinions Mining

Abstract

Identifying public policy agenda and relevant issues have served one of the crucial stages in public policy analysis. In addition to the traditional channels such as telephones and newspapers, the Internet has been emerging as the most challenging source of citizens’ complaints and comments. The existing practice and literature, however, appear insufficient to provide systematic investigation for conducting Internet public opinions analysis (IPOA). The present study reflects upon planning and implementing IPOA in a public agency via a series of interviews and field observation. The field experience contributes to the development of a step-by-step process to facilitate how public officials interact with consulting professionals and the IPOA service provider. Unlike transaction-oriented information systems, implementing IPOA is much similar to a decision support system that requires iterative communication and interpretation among three parties as specified above. Moreover, longitudinal volume and sentiment analyses have effectively provided fundamental insight. Nevertheless, the IPOA results appear to have potential limitation while the policy makers aspire to dig into the events correspondence and in-depth contents related to public attitudes and arguments concerning the policy examined.