A Technological Perspective on Information Cascades via Social Learning

• Main Authors: Fernando Rosas, Jen-Hao Hsiao, Kwang-Cheng Chen
• Format: Article
• Language: English
• Published: IEEE 2017-01-01
• Series: IEEE Access
• Subjects: Data fusion; distributed signal processing; sensor networks; social networks; collective behaviour; social learning
• Online Access: https://ieeexplore.ieee.org/document/7886356/

Collective behavior in human society is attracting a lot of attention, particularly as the result of novel emergent phenomena associated with online social media and networks. In effect, although crowd wisdom and herding behavior have been well studied in social science, the rapid development of Internet computing and e-commerce brings further needs of in-depth comprehension of their consequences and impact from a technological perspective. Based on social learning, an analytical knowledge originated in social science, we re-examine the well-known phenomenon of information cascade, whereby rational agents can ignore personal knowledge to follow a predominant social behavior triggered by earlier decisions made by peers. Moreover, we look into the cascade behavior from a communication theoretic perspective, interpreting social learning as a distributed data processing scheme. This perspective enables the development of a novel framework, which allows a characterization of the conditions that trigger information cascades and trace their impact on the accuracy of the collective inference. Finally, potential applications and examples of information cascade have been presented under various cyber technological scenarios, illustrating the prolific interplay between communication technology and computational social science.