Lattice-based Rank Aggregation algorithm (LARA)

Speaker: David CHENG

Abstract

Top-K Queries are popularly studied in the database community. Also it is clear that the distributed database becomes more and more common. In the distributed database environment, the bottleneck of the performance lies on the data transport instead of CPU computation. Therefore minimizing the number of data access is important.

In this talk, I would introduce our method called Lattice-based Rank Aggregation algorithm (LARA) to solve this problem. LARA would outperform other common methods, such as Quick-Combine and No Random Access (NRA), in terms of data access while maintaining a satisfactory running time.