Reasoning about Sentiment and Knowledge Diffusion in Social Networks

FABIO R. GALLO; GERARDO I. SIMARI; M. VANINA MARTÍNEZ; NATALIA ABAD SANTOS; MARCELO A. FALAPPA

IEEE INTERNET COMPUTING

IEEE COMPUTER SOC

Lugar: Los Alamitos, CA, USA; Año: 2017 vol. 21 p. 8 - 17

1089-7801

The past decade has seen an impressive adoption of Internet technologies for what is now commonly known as social media?different kinds of platforms that have the objective of connecting people in different ways. The aggregate of several such platforms can be seen as a complex network, in which there are many different kinds of relations, as well as different degrees of strength with which each one holds; moreover, these relations are dynamic because they can change over time. In this context, understanding how sentiments?and knowledge in general?expressed by agents (like people, institutions, automated bots) influence their connections is of great interest in many domains, such as marketing, activism, politics, or communication of ideas in general. In this paper, we show how the Network Knowledge Base model allows to represent the result of integrating information from multiple social networks, and then explore how information flow can be handled in this setting via belief revision operators that work on local (agent-specific) knowledge bases. Finally, we present preliminary experimental results on real-world Twitter data showing that different agent types react differently to the same information?these results are the first step in an effort to develop tools that will allow to predict how agents behave as information flows in their social environment.