Science des réseaux sociaux et économiques


Objectif

Aims

The study of social networks accompanied the development of quantitative sociology, and it is now blossoming thanks to the joint contribution of multiple disciplines, including physics and applied mathematics. It is also developing within economics, shedding light on non-market interactions that affect trade, labor markets and the diffusion of innovations.

The goal of the course is to provide students with fundamental insight into network science and how it can be used in sociology, economics, management, politics and neighboring disciplines. We will present essential concepts and theories at the intersection of different disciplines, showing applications to empirical problems. There will be elements of research design, focusing on how network science can be successfully integrated into a social science research project, and we will give you confidence in using key analytical tools and techniques with real-world data.

All social science (and other) backgrounds are welcome, and students are not assumed to have any previous knowledge of network science. Familiarity with Python is helpful but not essential.

 

Learning Outcomes

On attending this course, you will be able to:

–        Demonstrate knowledge of the key principles, approaches and achievements of network science

–        Understand network data type, source and format

–        Compute and interpret network metrics to analyze network data

–        Visualize network data

–        Discuss similarities and differences between network science and classical social science

–        Develop network-oriented research questions

–        Design data analysis approaches for network data

–        Apply basic modelling principles for network data

–        Distinguish small and large networks and use appropriate techniques for each type

 

Learning and teaching activities

We will meet eight times over January-March for three-hour sessions. Each session will consist of two parts, a lecture to present and discuss the main theories, and a tutorial to provide you with hands-on experience in network analysis and data visualization.

It is recommended that students bring their laptop with them. The proposed exercises will be done in Python.

There is also an important component of self-study with structured materials provided in class (see below).

 

Assessment

Take-home exercises (60%) and final project (40%).

Every week except holidays and the last session, you will be assigned a take-home exercise to consolidate your understanding of the ideas presented in class, to apply them in practice or to expand your knowledge of them. Some of these exercises are to be done individually, others in small groups. Each of these exercises will count 10% of your final mark. They are to be handed in electronically, the day before the start of the next class. In marking these exercises, we will value your effort, punctuality and seriousness more than your actual results.

At the end of the course you will be asked to submit a mini-research project (to be done in small groups) applying network science to a social-science topic of your choice. We will guide you toward the definition of your topic throughout the course, and the exercises are also designed to feed into your final project – in the sense that you will be able to gradually build on their outcomes. The project counts 40% and will be presented as a poster in a dedicated session after the end of the course. The final project is marked according to five criteria, with max 8% marks for each of them: 1/Originality of the topic, 2/Suitability of the interpretation, 3/Quality of analysis, 4/Methods, and 5/Presentation.

All submissions can be in English or French

Plan

Session Plan

An optional, preliminary session of 90 minutes on “Introduction to Python” is available for students with no previous familiarity with it. It will take place before the start of the course.

 

Module 1, sessions 1-2 (6 hours, PT): Introduction, notions and basic metrics

This module provides a general introduction to social network science and familiarizes students with basic concepts and analytical techniques for describing networks.

Sessions:

– S1: Introduction to the course; relationships and social structures; data formatting and storage; basic metrics and visualizations

– S2: Centrality measures, substructures, connectivity indicators and distance

 

Module 2, sessions 3-4 (6 hours, FG): Statistical approach to network science and theoretical models

This module first introduces the statistical approach to large network analysis. Using the notion of statistical ensables it consequently covers several network models: random graphs, small world networks, Barabasi-Albert graphs, generation of networks with attributes (such as homophily and geographical distance).

Sessions:

– S3: Statistical representation of large graphs. Random graphs, small world networks

– S4: BA graphs, generation of networks with attributes

 

 

Module 3, sessions 5-6 (6 hours, FG): Examples of model applications

This module covers applied examples of weighted networks (co-occurrence of hashtags in Twitter, collaboration networks, airport networks, trade networks…).

Sessions:

– S5: Weighted and multi graphs, typical measures, co-occurrence networks

– S6: Collaboration networks, airport networks, world trade network

 

Module 4, sessions 7-8 (6 hours, PT): Economic networks

This module introduces the economic and sociological literature that has used concepts, metrics and models from network analysis to shed light on substantive issues. It also introduces econometric models of network formation and dynamics, showing examples of applications.

 Sessions:

– S7: Social networks in organizations, trade and labor markets

– S8: Econometric models for network data with examples of applications

 

A final, 90 minutes session is organized after the end of the course for students’ poster presentations. Attendance to this session is compulsory.

Références

Barabási, Albert L. (2016). Network Science. Cambridge University Press.

Caldarelli, G. & Chessa, A. (2016). Data Science and Complex Networks: Real Cases Studies with Python. Oxford University Press.

Jackson, Matthew O. (2010). Social and Economic Networks. Princeton University Press.

Pastor-Satorras, R. & Vespignani, A. (2007). Evolution and Structure of the Internet: A Statistical Physics Approach. Cambridge University Press.

Robins, Garry (2015). Doing Social Networks Research: Network Research Design for Social Scientists. Sage.

 

Other readings (particularly research articles) will be proposed to students during sessions.