Disciplina Discipline MAE5908
Estatística de Redes Sociais

Social Network Statistics

Área de Concentração: 45133

Concentration area: 45133

Criação: 23/06/2020

Creation: 23/06/2020

Ativação: 23/06/2020

Activation: 23/06/2020

Nr. de Créditos: 8

Credits: 8

Carga Horária:

Workload:

Teórica

(por semana)

Theory

(weekly)

Prática

(por semana)

Practice

(weekly)

Estudos

(por semana)

Study

(weekly)

Duração Duration Total Total
4 2 4 12 semanas 12 weeks 120 horas 120 hours

Docente Responsável:

Professor:

Jefferson Antonio Galves

Objetivos:

Fornecer um quadro estatístico para analisar a transmissão da informação e a evolução do conjunto de opiniões e a formação de consenso em redes sociais.

Objectives:

To provide a statistical framework to model and analyze the transmission of information and the time evolution of opinion-based clusters in social networks.

Justificativa:

O curso discutirá questões estatísticas, probabilísticas, computacionais e de ciência de dados necessárias para modelar os processos de formação de redes sociais e identificar em tempo real como a informação se propaga nessas redes. Recentes análises estatísticas e sociológicas indicam que a propagação de informação e a formação de consenso em redes sociais seguem padrões novos, com características que parecem ser distintas daquelas observadas em processos sociais e de epidemia já repertoriados. Um exemplo disso é o fenômeno observado empiricamente de mudança de consenso em tempos curtíssimos em redes sociais. Modelar fenômenos desse tipo constituem um importante desafio científico, exigindo competências multidisciplinares. O estudo de redes sociais, além de exigir a construção de novos modelos probabilísticos, também apresenta grandes desafios para os estatísticos e os cientistas de dados, dado o caráter necessariamente parcial das amostras que podem na prática ser coletadas. Este curso abordará essas questões de um ponto de vista interdisciplinar, discutindo desde as ferramentas computacionais que podem ser usadas para o levantamento de dados nas redes sociais, até as questões de fundo que a análise de redes sociais coloca para a Teoria Estatística e para a Ciência de Dados.

Rationale:

The course will discuss statistical, probabilistic, computational and data-science issues required to model the processes of constitution of social networks and to identify how information spreads throughout these networks. Recent statistical and sociological analyses indicate that the spread of information and the formation of consensus, polarization, and clusters of opinion in social networks appear to differ from already-described social and epidemic processes. An illustration is the empirically observed phenomenon of fast consensus change in social networks. Modeling such phenomena is an important scientific challenge requiring multidisciplinary skills. The study of social networks not only requires the construction of new probabilistic models, but also presents major challenges for statisticians and data scientists, given the incomplete and partial nature of the samples that can be collected. This course will address these issues from an interdisciplinary point of view, discussing from the computational tools that can be used to the collection of data on social networks, up to the background questions that the analysis of social networks poses for Statistical Theory and Data Science.

Conteúdo:

1. Elementos básicos de teoria de grafos. Grafos aleatórios. O grafo de Erdos-Rényi. Transição de fase em gratos aleatórios. O modelo pequeno-mundo de Watts e Strogatz. Grafos com graus distribuídos segundo uma lei de potência. 2. Redes de opiniões evoluindo ao longo do tempo. O modelo clássico do votante, com e sem campo externo e robôs. Sistemas de votantes com memória de alcance variável interagindo entre si 3. Formação e evolução temporal de redes sociais. O modelo de Barábasi e Albert. Processos estocásticos assumindo valores num conjunto de grafos marcados. 4. Detecção de comunidades em redes sociais. O modelo de grafo de adesão a comunidades (community affiliation-graph model). Seleção estatística de modelos como ferramenta para detectar comunidades. Identificação de grafos de interação em sistemas de processos com memória de alcance variável interagindo entre si. 5. Comunidades homogêneas em redes sociais. Formação rápida de consenso em redes sociais. Identificação de campanhas de opinião a partir de amostras parciais.

Content:

1. Basic elements of graph theory. Random graphs. The graph of Erdos-Rényi. Phase transition in random graphs. The small-world model of Watts and Strogatz. Graphs with degrees distributed according to a power law. 2. Opinion networks evolving over time. The classical voter model with and without external field and robots. Interacting systems with variable range memory voters. 3. Formation and temporal evolution of social networks. The Barábasi and Albert model. Stochastic processes assuming values on a set of marked graphs. 4. Detection of communities on social networks. The community affiliation graph graph model. Statistical selection of models as a tool to detect communities. 5. Homogeneous communities on social networks. Rapid consensus building on social networks. Identification of opinion campaigns based on partial samples.

Forma de Avaliação:

Método/Critério:Aulas e exercícios. / O aproveitamento dos alunos será feito através da média ponderada de trabalhos, seminários, listas de exercícios e provas.

Type of Assessment:

The use of students will be done throught the weighted average of works, seminars, lists of exercises and tests.

Observação:

Bibliografia:

1. Brody, D.C. Modelling election dynamics and the impact of disinformation. Info. Geo. 2, 209–230 (2019). https://link.springer.com/article/10.1007%2Fs41884-019-00021-2 2. Collins, D., Efford, C., Elliott, J., Farrelly, P., Hart, S. Knight, J., Lucas, I. C., OʼHara, B., Pow, R., Stevens, J., Watling, G.: House of Commons Digital, Culture, Media and Sport Committee: Disinformation and ‘fake newsʼ: Final Report, Eighth Report of Session2017–2019. https://publications.parliament.uk/pa/cm201719/cmselect/cmcumeds/1791/1791.pdf 3. Duarte, A., Galves, A., Loecherbach, E., Ost, G.: Estimating the interaction graph of stochastic neural dynamics, Bernoulli 25(1), 2019,771–792 https://doi.org/10.3150/17-BEJ1006 4. R.A. Holley e T. M Liggett, Ergodic theorems for interacting infinite systems and the voter model, Annals of Probability, vol. 3, pag. 643-663, 1975 (https://www.jstor.org/stable/2959329?seq=1#metadata_info_tab_contents 5. Liggett, T.M. (1999). Stochastic interacting systems: contact, voter and exclusion processes. Springer. ISBN 3-540-65995-1. 6. Milgram, S.: The small world problem. Psychology Today 1(May):61-67 (1967). 7. Newman, M.: Networks, Oxford University Press, Second Edition (2018) DOI: 10.1093/oso/9780198805090.001.0001. 8. Travers, J., S Milgram, S,: An experimental study of the “small world” problem Sociometry, no 32(1969) 9. Vosoughi, S., Roy, D., Aral, S.: The spread of true and false news online. Science, Vol. 359, Issue 6380, pp. 1146-1151 (2018). https://science.sciencemag.org/content/359/6380/1146. 10. Watts, D. J., Strogatz, S. H.: Collective dynamics of small-world net works. Nature 393,440–442 (1998). 11. J. Yang and J. Leskovec, Community-Affiliation graph model for overlapping network community detection, 2012 IEEE 12th International Conference on Data Mining, Brussels, 2012, pp. 1170-1175. 12. Yang, L. Zhou, Z. Jin and J. Yang, "Meta path-based Information entropy for modeling social Influence in heterogeneous Information networks," 2019 20th IEEE International Conference on Mobile Data Management (MDM), Hong Kong, Hong Kong, 2019, pp. 557-562.

Bibliography:

1. Brody, D.C. Modelling election dynamics and the impact of disinformation. Info. Geo. 2, 209–230 (2019). https://link.springer.com/article/10.1007%2Fs41884-019-00021-2 2. Collins, D., Efford, C., Elliott, J., Farrelly, P., Hart, S. Knight, J., Lucas, I. C., OʼHara, B., Pow, R., Stevens, J., Watling, G.: House of Commons Digital, Culture, Media and Sport Committee: Disinformation and ‘fake newsʼ: Final Report, Eighth Report of Session2017–2019. https://publications.parliament.uk/pa/cm201719/cmselect/cmcumeds/1791/1791.pdf 3. Duarte, A., Galves, A., Loecherbach, E., Ost, G.: Estimating the interaction graph of stochastic neural dynamics, Bernoulli 25(1), 2019,771–792 https://doi.org/10.3150/17-BEJ1006 4. R.A. Holley e T. M Liggett, Ergodic theorems for interacting infinite systems and the voter model, Annals of Probability, vol. 3, pag. 643-663, 1975 (https://www.jstor.org/stable/2959329?seq=1#metadata_info_tab_contents 5. Liggett, T.M. (1999). Stochastic interacting systems: contact, voter and exclusion processes. Springer. ISBN 3-540-65995-1. 6. Milgram, S.: The small world problem. Psychology Today 1(May):61-67 (1967). 7. Newman, M.: Networks, Oxford University Press, Second Edition (2018) DOI: 10.1093/oso/9780198805090.001.0001. 8. Travers, J., S Milgram, S,: An experimental study of the “small world” problem Sociometry, no 32(1969) 9. Vosoughi, S., Roy, D., Aral, S.: The spread of true and false news online. Science, Vol. 359, Issue 6380, pp. 1146-1151 (2018). https://science.sciencemag.org/content/359/6380/1146. 10. Watts, D. J., Strogatz, S. H.: Collective dynamics of small-world net works. Nature 393,440–442 (1998). 11. J. Yang and J. Leskovec, Community-Affiliation graph model for overlapping network community detection, 2012 IEEE 12th International Conference on Data Mining, Brussels, 2012, pp. 1170-1175. 12. Yang, L. Zhou, Z. Jin and J. Yang, "Meta path-based Information entropy for modeling social Influence in heterogeneous Information networks," 2019 20th IEEE International Conference on Mobile Data Management (MDM), Hong Kong, Hong Kong, 2019, pp. 557-562.