MS&E 189 Lecture Notes - Lecture 4: Social Network, Matrix Multiplication
14 Jan 2017
School
Department
Course
Professor
Document Summary
Social network data document actors and relations between them. Conventional data describes actors and their own attributes. But networks are not interested in attributes. Data collection: actors: populations, boundaries, sampling, relations: content, direction, strength. N nodes representing actors and e edges representing relations. N={a, b, c, d, e} are nodes in accompanying graph and. E = {(a,d), (c, d), (b, e), (c, e), (d, e)} Requires that each node and edge is used only once. A graph is connected if for every path of nodes, there is a path between them. Matrix is mathematical representation of a graph as a two dimensional table. More efficient way to store network data when there are many actors and many relations. Certain mathematical operations such as addition and multiplication are allowed and useful as long as matrices are conformable. Asymmetric data: row represent sender and columns the target or relationship. Addition/subtraction: reflect the shift from simplex to multiplex relations.
