Relevant Work  Motivation  Some research areas might be very large (e.g., index structures, compression)  How do you find relevant scientific papers/thesis via multiple channels  By Venue/Year  Blind reviewing Scientific Reading [Graham Cormode: How NOT to review a paper: the tools and techniques of the adversarial reviewer. SIGMOD Rec. 37(4) 2008] This paper leaves many questions unanswered particularly poor. We use index structures like b-trees, tries, and hash tables; as well as compression techniques like run-length encoding, dictionary encoding, and null suppression. A woman without her

crucial to fit data in memory  automatic lossless compression  Sub-problems: #rows>>#cols, column correlation, column characteristics  column-wise compression w/ heterogeneous encoding formats Time Prog- Linear Algebra  Data characteristics inspired overall design of encoding schemes  Initially slow compression  dedicated sampling schemes and estimators  Initially slow compressed operations  cache-conscious it's an interesting problem General-purpose, heavy- and lightweight compression techniques struggle to achieve both good compression ratios and fast decompression speed to enable block-wise uncompressed

inclusion of header files). It makes testing much more difficult. Macros can do things these other techniques cannot, and you do see them in the codebase, especially in the lower-level libraries. And some metaprogramming Avoid complicated template programming. Template metaprogramming refers to a family of techniques that exploit the fact that the C++ template instantiation mechanism is Turing complete and can like Google Test, std::tuple, std::function, and Boost.Spirit would be impossible without it. The techniques used in template metaprogramming are often obscure to anyone but language experts. Code that uses

or lambda. Pros: Simple list comprehensions can be clearer and simpler than other list creation techniques. Generator expressions can be very efficient, since they avoid the creation of a list entirely actually run the file’s destructor. Different Python implementations use different memory management techniques, such as delayed Garbage Collection, which may increase the object’s lifetime arbitrarily and indefinitely

characteristics  Systematic evaluation w/ datasize, sparsity, etc  Inappropriate for certain topics: compression, ML accuracy  “Real” Data Repositories  Wide selection of available datasets w/ different characteristics