Cost Function Learning in Memorized Social Networks with Cognitive Behavioral Asymmetry

TL;DR

This paper introduces a novel social information diffusion model considering cognitive biases and proposes M$^{3}$IRL, a maximum-entropy inverse reinforcement learning framework, to learn individual cost functions from social media data.

Contribution

It presents a new social diffusion model incorporating asymmetric cognitive biases and a unique IRL framework that does not rely on MDP assumptions and requires only a single trajectory sample.

Findings

The social model accurately captures influence dynamics with cognitive biases.

M$^{3}$IRL effectively learns cost functions from limited social media data.

The approach outperforms traditional IRL methods in this context.

Abstract

This paper investigates the cost function learning in social information networks, wherein the influence of humans' memory on information consumption is explicitly taken into account. We first propose a model for social information-diffusion dynamics with a focus on systematic modeling of asymmetric cognitive bias, represented by confirmation bias and novelty bias. Building on the proposed social model, we then propose the M$^{3}$IRL: a model and maximum-entropy based inverse reinforcement learning framework for learning the cost functions of target individuals in the memorized social networks. Compared with the existing Bayesian IRL, maximum entropy IRL, relative entropy IRL and maximum causal entropy IRL, the characteristics of M$^{3}$IRL are significantly different here: no dependency on the Markov Decision Process principle, the need of only a single finite-time trajectory sample, and bounded decision variables. Finally, the effectiveness of the proposed social information-diffusion model and the M$^{3}$IRL algorithm are validated by the online social media data.