Back to search

Athens University Economics and Business (AUEB)

Combining relational and stream data for rea l- time analytics

Abstract

dc:description

Nowadays, more and more organizations realize the importance of analyzing available data. While most of the times data is stored, over the last years there is a growing amount of stream data. Such data arrives on-line from multiple sources in a continuous, rapid and time-varying fashion. Currently, most stream management applications and systems exploit stream data with the objective to answer monitoring queries. However, the real potential of stream data lies in the possibility to capture new types of information in (near) real-time and support decisions. To support this kind of analysis we need analytics queries that can support multiple and correlated stream aggregates over stream data coming from multiple and heterogeneous stream sources. Moreover a wide range of analytics applications need to combine already available data (e.g. stored data) and stream data to empower business with (near) real-time insights that can be used for improved decision making. As relational databases are extremely widespread our research focuses on how relation data can support relational-stream analytics applications. Overall, this thesis provides query formulation methods and tools that combine relational and stream data to support (near) real-time data analysis. In this thesis we introduce stream variables to support analytics over stream data. This kind of analytics queries can contain multiple stream aggregates, correlated stream aggregates and use data from multiple and heterogeneous stream sources. We provide SQL language extensions to support this kind of queries. Moreover we provide a spreadsheet-like approach to perform stream analytics. The intuition is that stream queries can by defined in a column-by-column fashion. The columns can contain either relational data or stream aggregates. The thesis studies how to extend current Relational Database Management Systems (RDBMSs) to handle stream data for (near) real-time decision making. We present a relational-based integration framework that sits atop any RDBMS and mix RDBMS’ data and stream aggregates managed by different stream systems. A SQL extension is provided to define relational-stream views and an API is developed to carry out the required communication between the relational and the stream systems. The proposed framework can serve as a standard for relational-stream interoperability.

Degree

thesis:*
Grantor dc:publisher
Athens University Economics and Business (AUEB)
Year dc:date
2013

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Sotiropoulos, Ioannis
  • Σωτηρόπουλος, Ιωάννης

Subjects

dc:subject × 6

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/43164

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Sotiropoulos, Ioannis; Σωτηρόπουλος, Ιωάννης. Combining relational and stream data for rea l- time analytics. Athens University Economics and Business (AUEB), 2013. http://hdl.handle.net/10442/hedi/43164