Abstract

Expert systems (ESs) are computer programs that use artificial intelligence strategies to accomplish sophisticated tasks once deemed possible only for human experts. This entry provides an overview of ways in which ESs have been used to facilitate social science research. It begins by outlining the architecture of ESs, then identifies common representation and reasoning strategies, and shows how those lead to a number of strengths as well as limitations. Next, this entry examines how ESs have been applied to key research tasks including designing research, expressing theories, inducing theory from data, coding, simulations, and data analysis. A few exemplary areas of ongoing application of ESs to research are noted. The entry concludes by characterizing the evolution of ESs over time. Modern intelligent systems have overcome most of the earlier limitations of ESs, integrating multiple intelligent systems to address a much broader scope of problems with more natural interfaces and more powerful reasoning systems. However, modern systems raise new concerns over lack of transparency, limitations due to the social construction of knowledge, and difficulties of research.