Sunday, October 12, 2014

Стартовал онлайн-курс "Social Network Analysis"

Совсем недавно на Coursera начался онлайн-курс Social Network Analysis, посвященный анализу социальных сетей, которой многим может показаться интересным и полезным, так как данная тема сейчас очень популярна и, кстати, хочу напомнить, что я какое-то время назад публиковал пост про встречу, посвященную тематике Data Science, на которой был представлен интересный доклад от Александра Семёнова по теме анализа социальных сетей: "SNA во времена Big Data: Насколько социален ваш анализ социальных сетей"?

Основным инструктором на данном курсе является Lada Adamic (Associate Professor, University of Michigan) из University of Washington. Курс продлится 9 недель с планируемой нагрузкой 5-10 часов в неделю.

Вступительное видео от создателей курса



Программа курса

  • Week 1: What are networks and what use is it to study them?
    Concepts: nodes, edges, adjacency matrix, one and two-mode networks, node degree
    Activity: Upload a social network (e.g. your Facebook social network into Gephi and visualize it ).
  • Week 2: Random network models: Erdos-Renyi and Barabasi-Albert
    Concepts: connected components, giant component, average shortest path, diameter, breadth-first search, preferential attachment
    Activities: Create random networks, calculate component distribution, average shortest path, evaluate impact of structure on ability of information to diffuse
  • Week 3: Network centrality
    Concepts: betweenness, closeness, eigenvector centrality (+ PageRank), network centralization
    Activities: calculate and interpret node centrality for real-world networks (your Facebook graph, the Enron corporate email network, Twitter networks, etc.)
  • Week 4: Community
    Concepts: clustering, community structure, modularity, overlapping communities
    Activities: detect and interpret disjoint and overlapping communities in a variety of networks (scientific collaborations, political blogs, cooking ingredients, etc.)
  • Week 5: Small world network models, optimization, strategic network formation and search
    Concepts: small worlds, geographic networks, decentralized search
    Activity: Evaluate whether several real-world networks exhibit small world properties, simulate decentralized search on different topologies, evaluate effect of small-world topology on information diffusion.
  • Week 6: Contagion, opinion formation, coordination and cooperation
    Concepts: simple contagion, threshold models, opinion formation
    Activity: Evaluate via simulation the impact of network structure on the above processes
  • Week 7: Cool and unusual applications of SNA
    Hidalgo et al. : Predicting economic development using product space networks (which countries produce which products)
    Ahn et al., and Teng et al.: Learning about cooking from ingredient and flavor networks
    Lusseau et al.: Social networks of dolphins
    Activity: hands-on exploration of these networks using concepts learned earlier in the course
  • Week 8: SNA and online social networks
    Concepts: how services such as Facebook, LinkedIn, Twitter, CouchSurfing, etc. are using SNA to understand their users and improve their functionality
    Activity: read recent research by and based on these services and learn how SNA concepts were applied

Более детальное описание курса можно найти на сайте курса на Coursera. Также там можно найти список рекомендуемой литературы.

No comments:

Post a Comment