Um estudo sobre alternativas de representação de dados temporais em bancos de dados relacionais

Content Provider	Semantic Scholar
Author	Cassol, Tiago Sperb
Copyright Year	2012
Abstract	Temporal information is present on a wide range of applications. Almost every application has at least one field that contains te mporal data like dates or timestamps. However, traditional databases don’t have a compreh ensive support to storage and query this kind of data efficiently, and DBMS with native support for temporal data are rarely available to system developers. Most of the time, regular databases are used to sto re application data and when temporal data is needed, it is handled using the po or support offered by standard relational DBMS. That said, the database designer m ust rely on good schema design so that the natural difficulty faced when dealing with temporal data on standard relational DBMS can be minimized. While some design choices may seem obvious, others ar difficult to evaluate just by looking at them, therefore needing experimentati on prior to being applied or not. For example, in several cases it might be difficult to measure how much will a specific design choice affect the disk space consumption, an d how much will this same design choice affect overall performance. This kind of inf ormation is needed so that the database designer will be able to determine if, for example, the increased disk space consumption generated by a given choice is acceptab l because of the performance enhancement it gives. The problem is that there is no study that analyses th design choices available, analyzing them through concrete data. Even when it is easy to see which of two design choices perform better in a given criterion, it is hard to see how better the better choice does, and if any other side-effect it has is accept able. Having concrete data to support this kind of decision allows the database designer to make the choices that suits his application’s context best. The objective of this work is to analyze several co mmon design choices to represent and handle different kinds of temporal data on stan dard SQL DBMS, providing guidance on which alternative suits best each situa tion where temporal data is required. Concrete data about each of the studied alternative s ar generated and analyzed, and conclusions are drawn from them.
File Format	PDF HTM / HTML
Alternate Webpage(s)	https://lume.ufrgs.br/bitstream/handle/10183/67849/000874007.pdf;sequence=1
Language	English
Access Restriction	Open
Content Type	Text
Resource Type	Article