Midterm Exam
Monday, Mar 19th, 1:50pm - 3:05pm, Stuart 104
The exam will be 75 minutes long. It is closed book and closed notes. Some questions will be multiple-choice. Topics covered will be:
- Datalog
- Write datalog queries
- Translate from datalog to SQL or relational algebra
- Constraints
- Write constraints as first-order logical formulas.
- Translate constraints into first-order logical formulas. For example, given a functional dependency translate it into its first-order representation.
- Translate constraints into denial constraints
- Write denial constraints
- Query Equivalence and Containment
- Find containment mappings between queries to determine their containment relationships.
- Create Schema Matchings
- Determine schema matches based on simple schema-based matches such as edit distance with threshold or Jacard with threshold.
- Virtual Data Integration
- Rewrite a query using a set of views using the inverse rules or bucket algorithm.
The exam is available here. You can find the solutions here.
Final Exam
Thursday, May 3rd, 10:30am-12:30pm, Stuart 104
The exam will be 120 minutes long. It is closed book and closed notes. Questions will be both multiple-choice and full text questions. These topics have already been covered in the homework assignments. Topics covered will be:
- Data Warehousing
- Write SQL queries using the generalized grouping and windowed aggregation (analytical function) features we have discussed in class.
- Answer basic multiple choice questions about the multidimensional datamodel and its abstract operations (drill-down, roll-up, ...).
- Big Data Analytics
- Mostly multiple choice questions and short text questions about the topics we have covered such as
- Fault tolerance and Load balancing.
- Distributed filesystems and specifically the implementation of this concept in HDFS.
- The MapReduce programming model and the key-value datamodel.
- The implementation of MapReduce in Hadoop, how the map and reduce phases operate, how the system achieve fault tolerance and load balancing.
- Provenance
- Answer multiple choice questions about provenance models, e.g., insensitivity to query rewrite
- Know how to determine the provenance of a simple query according to one of the following provenance models.
- The 3 variants of Why-provenance: the set of witnesses, Why-provenance, and minimal Why-provenance.
- How-provenance (provenance polynomials).
The exam is available here. You can find the solutions here.