Current Trends in Database Models

Current Trends in Database Models

The evolution of database system models over the last 30 years has influenced many entities from the fields of business communications to marketing systems. These database system models have evolved from simple systems that manage simple flat files to complex software engines designed to handle enormous amounts of data. Some current database trends reveal a response to two fast growing technology areas: web-based applications and database systems, which refers to database systems used for data mining. These database systems manage enormous amounts of information and data companies and businesses can use to create information or knowledge through discovery of previously unknown correlations and associations between the data. Some current and emerging trends involve multimedia industry databases, columnar database model, object-oriented model, and schema free document oriented models.

Today there are "multimedia databases, distribution databases, document oriented databases and mobile and embedded databases" ("Current Trends in Databases," 2013, Slide 7). One database that has seen growth over the past few years has been in the multimedia industry. As the growth of digital images, music for MP3 players, and websites such as YouTube, and Goggle continue to grow the need for storage and processing of this data has more than quadrupled over the past few years. Web-based companies with the ability to give web users access to these databases to retrieve information and view their content is the key to any success. Individuals can perform key word searches that could include any of the following: the day of the photograph or video, artist name or song title, time or place a photo, or perhaps the name of the person that recorded it. Giving end users access to company databases can raise security concerns, but with proper protocols these issues rarely happen.

There was a time when a columnar database was contending for the top spot as the primary architecture for structuring databases. The increase in processing power, memory, and the lower cost of hardware, businesses were continuing to use row-based database systems. The columnar architecture fell to the wayside despite the performance advantages in large analytical queries that involve a few types of data elements (Raab, 2007). With pressures to decrease IT costs the columnar database may once again rise in popularity. One issue that needs to be addressed with current columnar databases is the load time. As a columnar database typically has to convert data from a row-based structure, load times are typically slow relative to the load times of a relational database. In addition, decompression (if the data is compressed) may contribute to load times lagging (Raab, 2007). Load time is one of the main reasons the conventional row-based architecture has been more popular than columnar offerings.

New techniques employed by the latest columnar database offerings can increase the speed of load times such that they are less of a concern (Monash Research, 2008). The vertical offering, for example, runs in hybrid mode, making data available immediately for queries once loaded into memory. In addition, the faster processor and increased memory of current systems tends to decrease load time for columnar databases to that comparable with conventional