Social Data and Networks
Date: TBD 2016 | Tuition: TBD | Continuing Education Units (CEUs): TBD
*This course has limited enrollment. Apply early to guarantee your spot.
Application Deadline »
Networks are ubiquitous in the modern era, be they social media networks such as Facebook and Twitter, transportation networks formed by aerial or ground routes, political networks observed through blogs and opinions, energy dispatch networks formed between end-users and generators, or meta-networks observed in ‘inter-dependent’ data such as World Wide Web or biological protein interaction.
We know a great deal about these networks. All things online are recorded, cheap sensors of all sorts are providing a wealth of information about operations, and experiments can be performed at a massive scale.
Understanding and utilizing such networks can help tremendously in making better societal decisions regarding public policies, business operations, financial market regulations, utility network operations, and advancing fundamental science.
In this course, we shall provide an in-depth, state-of-the-art analytic view of social networks based on behavioral models and statistics, with the goal of making use of these networks. The course will be primarily about concepts and fundamentals; for illustrative purposes appropriate case studies will be discussed. It will provide principled guidelines for utilizing social networks for operational purposes.
Fundamentals: Core concepts, understandings, and tools (30%)
Latest Developments: Recent advances and future trends (40%)
Industry Applications: Linking theory and real-world (30%)
Lecture: Delivery of material in a lecture format (80%)
Discussion or Groupwork: Participatory learning (10%)
Labs: Demonstrations, experiments, simulations (10%)
Introductory: Appropriate for a general audience (70%)
Specialized: Assumes experience in practice area or field (25%)
Advanced: In-depth explorations at the graduate level (5%)
The participants of this course will be able to:
- Utilize social data operationally for various decision making tasks, e.g. advertising and marketing, policy making, better serving customers, etc.
- Develop effective summary of social data in the form of meaningful models.
- Design scalable algorithms for social data processing.
- Utilize networks and graphical models.
- Obtain qualitative insights from dynamic network models.
Who Should Attend
Statistician/data scientists, marketing officers (CMOs), data analytic experts, researchers in social sciences, faculty in social sciences and marketing, and political bloggers.
Day One: Social Data, Networks, and Dynamics
9:00 – 10:30 am Social data and networks, models, and dynamics (M. Dahleh)
- What's available: private vs. public, e.g. Facebook, Twitter, Yelp, blogs, etc.
- What we can expect from them
• ‘Tipping point,’ successful marketing/political campaign and societal revolutions
- Why networked and importance of being networked
- Models of networks and dynamics over networks
10:30 – 11:00 am Coffee break
11:00 am – 12:30 pm Consensus, opinions, and networks (M. Dahleh)
- What is consensus? Play a simple game.
- When does it work? What do you converge to?
- Disagreement model
- Information spread, conductance and bottleneck
12:30 – 1:30 pm Lunch break
1:30 – 3:00 pm Learning in networks (D. Shah/M. Dahleh)
- Wisdom of the crowd
- Detection of source
- Rumor centrality, source estimator and importance of individuals
- Learning influence with temporal data
3:00 – 3:30 pm Coffee break
3:30 – 5:00 pm Centrality (M. Dahleh)
- Network centrality: quantitative network measure
• Operational utility
• Examples: Rumor, Page Rank, Bonacich, Degree
- How to compute?
- A case study: social pricing/advertising using network centrality
5:00 – 7:00 pm Open session and social mixer
Day Two: Social Data, Networks, and Decisions
9:00 – 10:30 am Communities in network. (D. Shah)
- What are communities? Stochastic Blockade model
- Why should we detect them (how can they help)?
- How to detect them?
• Modularity centrality, K-means, and Leader-Follower
10:30 – 11:00 am Coffee break
11:00 am – 12:30 pm Polling, ranking, and policy making (D. Shah)
- Why do we rank? To make decisions.
- How to seek people’s input? Through comparisons
- How to aggregate opinions? Impossible and then not impossible
- Rank centrality, Bradley-Terry-Luce model of judgment
12:30 – 1:30 pm Lunch break
1:30 – 3:00 pm Recommendation (D. Shah)
- Importance of personalization: Netflix, Amazon, Yelp, and Health
- A simplistic view: ranking
- A Bayesian view: collaborative filtering
- Network view: Collaborative centrality
3:00 – 3:30 pm Coffee break
3:30 – 5:00 pm Crowd-Sourcing (D. Shah)
- The power of the crowd – more than computers
- Budget optimal crowd campaigns
- Crowd opinion aggregation and crowd centrality
- A case study: Using Amazon’s Mechanical Turk
5:00 – 6:00 pm Open session
Course schedule and registration times
Class runs 9:00 am - 5:00 pm on both days, followed by an open session that runs until 6:00 -7:00 pm.
it manager, medtronic
"The instructors were very open to further discussions after the course ended, and they seemed genuinely interested in the business problems the participants were facing."
"I plan to start immediately sharing the knowledge I gained in the course, and this will bring a completely new perspective to what we have been doing so far."
product manager digital, editrice la stampa
"I’ll definitely benefit from the acquired knowledge, in particular regarding recommendation engines and network detection as I am working on those topics for marketing purposes in my company."
chief representative, nbso frankfurt/main
"The course was interesting; the opportunities that arise out of the presented content are very interesting. The rating of the course as such is therefore very positive. In my opinion, there is a high demand for this information."
principal network architect, cox communications
"We have two gifted instructors because they possess an in-depth knowledge in a timely and vexing topic and were able to deliver it an easy to follow approach."
About The Lecturers
Devavrat Shah is currently a Jamieson Career Development Associate Professor with the department of Electrical Engineering and Computer Science at MIT. He is a member of the Laboratory for Information and Decision Systems (LIDS) and Operations Research Center (ORC). His current research interests include social networks and statistical inference.
His work has been recognized through paper awards in machine learning, networking, and operations research; career prizes 2008 ACM Sigmetrics Rising Star award and 2010 INFORMS Erlang Prize; and recent press releases in the New York Times (on recommendation systems) and Forbes (on predicting trends in Twitter).
Munther Dahleh is the current director of the Engineering Systems Division at MIT. Previously, he was the Associate Department Head of the department of Electrical Engineering and Computer Science.
Dr. Dahleh is interested in problems at the interface of robust decisions, filtering, information theory, networks, and computation. In addition to methodology development, he is interested in the application of distributed control in the future electric grid and the future transportation system with particular emphasis in the management of systemic risk. He is also interested in various problems in social networks including information propagation over complex networks, global games and application to revolution models, and various connections to health care.
This course takes place on the MIT campus in Cambridge, Massachusetts. We can also offer this course for groups of employees at your location. Please complete the Custom Programs request form for further details.
Links & Resources
- Devavrat Shah on Twitter trends Shah discusses how he uses his machine-learning algorithm to predict trends hours in advance of recognition by Twitter.
- “Resilience of Dynamical Transportation Networks" Dr. Munther A. Dahleh (ICINCO and SIMULTECH 2012)
- Predicting what topics will trend on Twitter A new algorithm predicts which Twitter topics will trend hours in advance and offers a new technique for analyzing data that fluctuate over time.
- Dahleh appointed to new roles Leading ESD and planning for a potential MIT-wide effort in the areas of socio-technical systems, information and decision systems, and statistics.