Abstract:By "social learning," in this article we refer to mechanisms for opinion formation and decision making over graphs and the study of how agents' decisions evolve dynamically through interactions with neighbors and the environment. The study of social learning strategies is critical for at least two reasons. On one hand, it allows for a deeper understanding of the fundamental cognitive mechanisms that enable opinion formation over networks and the propagation of information or misinformation over them. On the other hand, these same learning strategies are effective for decision making by networked agents under challenging conditions, such as highly dynamic environments, nonstationary models and data, untruthful or malicious agents, sparsely connected graphs, and constrained communication. The article presents a unifying framework that covers several cases of interest, such as single-agent Bayesian learning, multiagent non-Bayesian learning, adaptive social learning, social machine learning, partial information sharing, influence discovery, and many others. The presentation highlights important limitations of the traditional social learning strategies. One limitation is the inability to track well drifting conditions. Traditional approaches lead to stubborn agents, which resist new states of information and are slow to react to changes in the environment, like an opinion that changes over time. Another limitation of the traditional strategies is that they assume perfect knowledge of the data models, which is seldom available in practice. The article illustrates recent advances that address these issues. We show how to endow multiagent networks with adaptation abilities and how to build social machine learning solutions that learn the necessary models directly from the data. These are fundamental steps toward the construction of socially intelligent networks, capable of exploiting cooperation and diversity across the agents to guarantee reliable learning performance under nonstationary, heterogeneous, and uncertain environments.

The wisdom_of_crowds: an efficient, philosophically-validated, social epistemological network profiling toolkit

Contributive Social Capital Extraction From Different Types of Online Data Sources

OCTOPUS: an Online Topic-Aware Influence Analysis System for Social Networks

When social influence promotes the wisdom of crowds

Wiser than the Wisest of Crowds: The Asch Effect Revisited under Friedkin-Johnsen Opinion Dynamics

Network dynamics of social influence in the wisdom of crowds

Socially Intelligent Networks: A framework for decision making over graphs

Learning from Crowds in the Presence of Schools of Thought.

Crowdsourced science: sociotechnical epistemology in the e-research paradigm

Leveraging Peer Centrality in the Design of Socially-Informed Peer-to-Peer Systems

Diversity of a User's Friend Circle in OSNs and Its Use for Profiling.

A Transfer Learning Based Framework Of Crowd-Selection On Twitter

Knowledge, Trust, Security and Covertness In Massively Distributed Social Platforms: An Epistemic Networks Approach

2 SocialImpact : Bring Order to Online Underground Social Dynamics

An Exploratory Data Analysis of the #Crowdfunding Network on Twitter

Efficient Crowd Exploration of Large Networks: The Case of Causal Attribution

Comparing Echo Chamber Detection Metrics: A Cross-modeling and Cross-platform Analysis of Twitter and Reddit

Ceka: A Tool for Mining the Wisdom of Crowds

Your most telling friends: Propagating latent ideological features on Twitter using neighborhood coherence

Social network analysis of Twitter interactions: a directed multilayer network approach

Finding the True Crowds: User Filtering in Microblogs