Welcome to the Capstone Project for Big Data! In this culminating project, you will build a big data ecosystem using tools and methods form the earlier courses in this specialization. You will analyze a data set simulating big data generated from a large number of users who are playing our imaginary game "Catch the Pink Flamingo". During the five week Capstone Project, you will walk through the typical big data science steps for acquiring, exploring, preparing, analyzing, and reporting. In the first two weeks, we will introduce you to the data set and guide you through some exploratory analysis using tools such as Splunk and Open Office. Then we will move into more challenging big data problems requiring the more advanced tools you have learned including KNIME, Spark's MLLib and Gephi. Finally, during the fifth and final week, we will show you how to bring it all together to create engaging and compelling reports and slide presentations. As a result of our collaboration with Splunk, a software company focus on analyzing machine-generated big data, learners with the top projects will be eligible to present to Splunk and meet Splunk recruiters and engineering leadership.
Simulating Big Data for an Online Game
This week we provide an overview of the Eglence, Inc. Pink Flamingo game, including various aspects of the data which the company has access to about the game and users and what we might be interested in finding out.
Acquiring, Exploring, and Preparing the Data
Next, we begin working with the simulated game data by exploring and preparing the data for ingestion into big data analytics applications.
Data Classification with KNIME
This week we do some data classification using KNIME.
Clustering with Spark
This week we do some clustering with Spark.
Graph Analytics of Simulated Chat Data With Neo4j
This week we apply what we learned from the 'Graph Analytics With Big Data' course to simulated chat data from Catch the Pink Flamingos using Neo4j. We analyze player chat behavior to find ways of improving the game.