CACM Research Highlight: Crowdsourcing Enumeration Queries

Each month the Communications of the ACM publishes one or two “Research Highlight” papers chosen from across the field of Computer Science.  These papers are intended to showcase cutting edge work in various sub-disciplines of CS research for the wider tech audience.  The January 2016 issue of CACM contains a paper by AMPLab researchers (and alums) Beth Trushkowsky (now a Professor at Harvey Mudd), Tim Kraska (now a Professor at Brown), Mike Franklin, and Purna Sarkar (now a Professor at UT Austin) entitled “Answering Enumeration Queries with the Crowd“.   This is an extended version of a paper from Beth’s Ph.D. research that won the Best Paper award at the 2013 ICDE (International Conference on Data Engineering) Conference.

The paper comes squarely from the “People” side of the “Algorithms, Machines and People” research agenda of AMPLab.   It looks at the question of how to use crowdsourcing platforms to collect lists of items such as a list of all the Ice Cream flavors in the world, or all the restaurants in the Bay Area that serve Poke  (it works for non-food related questions as well).   The paper raises questions about the semantics of such queries in a crowd environment, and provides methods to derive reliable termination conditions for such crowdsourced queries.   The paper is preceded by a nice one-page “Technical Perspective” summary of the article and its significance written by AMPLab crowdsourcing collaborator Tova Milo of Tel Aviv University.

This is the second AMPLab paper chosen as a CACM Research Highlight.   The paper “Quantifying Eventual Consistency with PBS”  by Peter Bailis, Shivaram Venkataraman, Mike Franklin, Joe Hellerstein, and Ion Stoica appeared as a Research Highlight in the August 2014 issue of the journal.