Question 1 :
How would you go about designing a very large, rapidly changing dimension in a multi-dimensional data warehouse? Describe the design and how it would work along with any potential issues you feel that effect performance?
Question 2 :
Suppose a retail business (e.g., Coca-Cola) was interested in monitoring online discussion about their product and their competitor's products. What approach would you take, what tools would you use in this approach, and what sort of insights would you try to extract from the data you obtained?
Question 3:
You want to identify communities in a graph. When would you use community detection techniques based on edges, and when would you use data clustering techniques based on similarity measures?