数据库领域目前的研究方向

概述了过去两年SIGMOD、ICDE和VLDB等顶级会议的研究成果。指出当前数据库研究的主要方向包括数据智能分析、分布式计算与系统优化以及AI技术在数据处理中的应用等方面。

(1)使用MapReduce来处理一些传统的问题

Efficient parallel set-similarity joins using MapReduce(SIGMOD 2010)

Scalable distributed non-negative matrix factorization is used for analyzing binary relationship data of web-scale

mapreduce(www 2010)

A comparison of join algorithms for log processing in MapReduce(SIGMOD 2010)

(2)MapReduce的优化(性能、索引、查询语言等),例如：

MapDupReducer : Detecting Near Duplicates over Massive(SIGMOD 2010)

Mapdupreducer: detecting near duplicates over massive datasets(ICDE)

The Performance of MapReduce: An Indepth Study(VLDB 2010)

On single-pass indexing with MapReduce(SIGIR 2009)

ASSET Queries: A Declarative Alternative to MapReduce (SIGMOD 2009)

MR-Scope: A Real-Time Tracing Tool for MapReduce (HPDC 2010)

A Platform for Scalable One-pass Analytics using MapReduce(SIGMOD 2011)

(3) MapReduce+PDBMS(NoSQL+SQL)

Integrating hadoop and parallel DBMs(SIGMOD 2010)

HadoopDB：基于MapReduce技术和数据库管理系统的混合体用于分析场景

Workloads(VLDB 2009)

A comparison of approaches to large-scale data analysis(ICDE 2009)

(4) 列存储以及行列混合存储机制

Read-Optimized Databases, In Depth(VLDB 2008)

ColumnStores vs. RowStores: How Different Are They Really?(SIGMOD 2008)

(5)支持特殊数据处理的专用数据库的研究

ArrayStore: An Efficient Storage Manager for Handling Complex Parallel Array Processing Tasks (SIGMOD 2011)

Overview of SciDB(SIGMOD 2010)

本文的部分文章列表源自[此处链接]（可直接点击访问），特致谢！