Intel’s BigDAWG project will break new ground in accessing data that spans multiple data models efficiently and effectively. We propose to study query optimization within this system. In particular, our work will investigate summarizing the data management needs of individual queries. We will do so by computing a query signature based on the user’s input text. We will also research using signatures to probe the space of possible engines for executing a given query. This probing will enable users to focus on “what” their queries will do by abstracting away the “how”.
|Effective start/end date||9/1/15 → 12/31/19|
- University Industry Research Corporation (Letter 1/29/16)