Social Network Analysis Sandbox
This is where @aguestuser will put problem sets and toy projects for learning Social Network Analysis as a side project for their work on LittleSis.
Texts
Currently Reading:
- Social Network Analysis by John Scott -- solid intro w/ history of field, treatment of major algorithms, written from perspective of a sociologist mapping corporate power trying to learn the math
- O'Reilly Python Data Science Handbook by Jake VanderPlas -- comprehensive overview of python for data science: including numpy, pandas, machine learning, linear algebra, etc.. Bonus: there's a version of the book published completely in IPython notebooks!
Might read
- Social Network Analysis: Methods and Applications by Stanley Wasserman -- heavy-on-the-math treatment of SNA methodologies, good to read after the Scott book?
- Analyzing the Social Web by Jennifer Goldbeck -- this one is all about mining social media platforms for social network data! we should totally learn how to do this for corporate targets! turn the weapons against their makers!
- Analyzing Social Networks by Borgotti, Everett, and Johnson -- this is a captivating book: written by researchers and software developers with a heavy emphasis on how to model graph analysis in computational algorithms. however the software-driven bits seem tightly coupled to an outdated, proprietary graph analysis software tool called UCINET that only runs on Windows. I discarded it for that reason, but maybe a closer look will reveal that the book is more translatable into our context than I ghouth?
- Understanding Social Networks: Theories, Concepts, and Findings by Charles Kadushin -- broad overview for non-mathmatical audience, seems redundant with other choices above, namely the Scott book, which seems better written
Tech
graph analysis libraries (python)
- Graph Tools super-performant, well-documented, expressive library that can model property graphs (ie: where data is stored on edges)
- SNAP the gold standard for graph analysis in the sciences, also super-perofrmant and well-documented, but does not model property graphs
distributed graph analysis (scala)
- Apache Spark -- distributed data analysis platform
- GraphX -- spark libary for graph analysis
- Spark Packages: Graph -- graph-related plugins for Spark
- Breeze scala equivalent of numpy
- ScalaNLP Suite -- includes breeze for linear algebra, epic for statistical parsing, and puck for GPU-powered parsing
- BigDL -- intel-made platform for deep learning on Spark
graph databases