Statistical physics concepts in modeling online emotional data

Datum

26.03.2014

Zeit

11:00 - 12:00

Sprecher

Julian Sienkiewicz

Zugehörigkeit

Warsaw University of Technology

Sprache

en

Hauptthema

Physik

Andere Themen

Physik

Host

Eduardo Altmann

Beschreibung

Social groups interacting via electronic media have recently gained particular interest. The creation and development of automatic sentiment analysis has made large scale emotion detection and analysis possible using text messages collected from the web. In the talk I show how two well-known statistical physics concepts, namely entropy and a modified version of the Ising model, can be used to track the evolution chronologically ordered emotional discussions in online media. First I analyze data from Internet relay chat (IRC) as well as from BBC Forum website and model the dialogues between chat participants by assuming that the driving force for the discussion is the entropy growth of emotional probability distribution. This process can be claimed to be responsible for a power-law distribution of the discussion lengths observed in the dialogues which is confirmed by numerical simulations. In the second part of the talk I introduce a growing one-dimensional quenched spin model that bases on asymmetrical one-side Ising interactions in the presence of external field. Numerical simulations and analytical calculations based on Markov chain theory show that when the external field is smaller than the exchange coupling constant J there is a nonmonotonous dependence of the mean magnetization on the temperature in a finite system. The crossover temperature Tc corresponding to the maximal magnetization decays with system size, approximately as the inverse of the Lambert W function. By performing tests on real data I conceive the model is in part suitable for a qualitative description of online emotional discussions arranged in a chronological order.

Letztmalig verändert: 26.03.2014, 08:44:32