A. A conceptual model of Multiple Dimensions of Identity 

   1. What might be some elements of a person's identity in a computer system? E.g. , what are roles people identify with such as  manager, developer, or player? 
   2. In what ways can the above roles overlap?

   3. Is there a core set of roles in your answer (1) that do not overlap? 

B. Identity management in cyberspace

   1. Explain how the media used for an Internet service, i.e., text, audio 
or video, could help either integrate or dissociate a person's identity 
for an Internet Service. 

C. Robust De-anonymization of Large Sparse Datasets

  1. Given a database D with records R, for which some of the columns altered,
say replacing the keys with random hash values, give a English language description
of what de-anonymization does. 

  2. Suppose you have a database D with a record set R, for which some of 
the columns have been changed to remove the identifying information, say 
replaced with a random hash value. Describe what the relationship of the 
columns and rows of the auxiliary database, Aux, to D. Generally, what
information does Aux need to have to de-anonymize D? 

   3. Explain equation 2 on page 3 in words. What does it mean for a 
database to be (e,d) sparse? That is, what is the intuition of sparsity? 

   4. What are the 3 main components of the authors' de-anonymization 
algorithm? 

   5. Why is sparsity the key feature needed for the authors' approach? What 
might happen to their de-anonymization algorithm if the data were dense, instead of sparse? 

6. Describe the Equation of Lemma 1 on page 6 in English text.