Desktop Site (Beta)

Introductory Video

Case Study in Travel Business - Understanding agent connections using NetworkX

Loading

Follow to receive video recommendations   a   A
Speaker: Are you the speaker?

Network analysis is getting more and more attention in BusinessIntelligence, people hope to get information out of the structure of anorganization or a communication network. In this talk, we use the hotelroom search requests from travel agents, including online publicwebsite, B2C, B2B and B2B2C, to build a relational network among them.By using this network as an example, we demonstrate how insights can beextract by studying network properties.In the first half of the talk, we will explain how the network is builtusing NetworkX, an open-source python library that is designed for thecreation, manipulation, and study of the structure, dynamics, andfunctions of complex networks. When 2 agents are making the same searchat the same time , a link ( or an “edge” in network analysts terms) ismade pointing form the initial searcher to the subsequent searcher.Using a list of these searches, a directed graph is built. We will alsodemonstrate how to pick the biggest connected component out form thegraph. In the second half, with the graphs created, we show howdifferent functions of NetworkX can be used to study the graphs. Bycompare the graph properties of our graph to the other popular networkgraphs, we can get the insight of how the network was created. Also bystudying the graphs, we can understand the behavior of the agents andcan even figure out which agents are acting as main hubs in the network.This talk is for people who are interested in network analysis and wouldlike to see how it can be used in a business case. Audiences with anylevel of python experience can learn some basic concept of networkanalysis work and how it can be applied to provide business insights.

Editors Note:

I would like to work with open source projects to create a branch of the tree with all of the best videos for your open source project. Please send me an email if you are interested.  

Comment On Twitter