In the mid 1970s IBM computer scientist E.F. Codd developed the relational model for database management (RDBMS). In the early 1980s the relational model became the preferred model for database management. Although, network or hierarchical databases are sometimes used instead because they are viewed by some as being more user friendly, RDBMS is still the most popular data management system (Fontecchino, 2015). It is most frequently used for data contained in logistical information, manufacturing information, financial records, and in personnel data for companies. It has withstood new developments in database management for over thirty years. RDBMS is favored by iconic tech companies such as Oracle, IBM, Microsoft and SAP for multiple reasons (Fontecchino, 2015).
Like all data management systems RDBMS stores, retrieves, and processes data in addition to reporting and controlling who views and changes data. This management system is known for its flexibility, specifically its ability to allow users to re-examine older data in newer ways. The main benefit of RDBMS is that it can organize the entered information and provides the user with values that are both reasonable and consistent with other data entered (Fontecchino, 2015.
RDBMS are very popular. However, they are not universally used and have their flaws. RDBMs are one of the more structured data management systems. Generally, RDBMS store data in rows from a fixed number of pre-defined fields. Data that comes from unknown fields or many different sources. will not display well in RDBMS and the system may have issues with outputting values for such data. As newer apps become more media-laden the accompanying data becomes less structured with fewer rules governing integrity. The lack of structure presents a problem in RDBMS because the unstructured data does not fit well into the rows that make up RDBMS (Susan, 2007). Those looking for an alternative to RDBMS may use an object-oriented database management system instead. Object oriented databases is table oriented, uses columns instead or rows and sorts data by object, class and inheritance (Susan 2007).
All companies are reliant on good data to preform. Consequences of bad data include higher rates of customer turnover, employee frustration, missed sales opportunities, and avoidable company expenses resulting from tasks needing to be repeated because of decisions informed by bad data (Susan, 2007). There are many factors that can affect the quality of data. Some of these include organization, validity of data domains, location consistency, integrity, accuracy and relevance. These factors will also influence what kind of data management system will work best for a company because, while every company needs data management system, not every data management system is appropriate for every company. RDBMS are best suited for companies with high levels of IT security and a larger IT force as RDBMS can be accessed by multiple users at the same time (Susan, 2007).
A company looking to store and analyze data that contains customer contact information or records from past employee reviews would do well with a RDBMS. Basically, any data that would be gathered by hand by having the user fill out a form with pre-labeled fields would work well in a RDBMS. A company would not want to use a RDBMS to manage open-ended, unorganized or data from media-heavy sources. More media generally means less data will go into the fields RDBMS need to sort data into rows and the unstructured data will alter the way RDBMS calculates values. If a company wanted to store something like, customer reviews of a product that would not want to use a RDBMS instead, they would want to store that data using a hierarchical data base management system instead so they can closely track the relationship between the variables such as the profile of the customer giving the review and the different facets of data from the review like comments or number ratings.
Many social media companies and app developers deal with a lot of unstructured data like pictures, videos, html links and other text with no definite origin or category (Unstructured Data: a big deal in big data, 2015). These companies struggle with large amounts of data that are difficult to structure and store in a RDBMS. Twitter is one such well known company. The social networking platforms deals in tweets, short 140 character messages that are often embedded with all different kinds of media. Although, unstructured data can be difficult to deal with it is sill highly valuable to companies because “can translate into connections and patterns that would otherwise be missed. Depending on the mission, those missed connections could result in real missed opportunities” (Unstructured Data: a big deal in big data, 2015, p.2). Twitter’s information is stored… Data is then organized and analyzed used a number of different tools. The first tool is a platform called Hadoop, an analytic tool made by scientists at Google where it was “designed to process, store and analyze petabytes and exabytes of distributed, unstructured and structured data” (Unstructured Data: a big deal in big data, 2015, p.3) Other database management strategies called ETL (Extract/Transform/Load) tools. Twitter is just one of many companies who use unorthodox tools to manage unstructured data.
- Fontecchino, M. (2015). Oracle the clear leader in $24 billion RDBMS market. TechTarget. Retrieved 29 December 2015, from http://itknowledgeexchange.techtarget.com/eye-on-oracle/oracle-the-clear-leader-in-24-billion-rdbms-market/
- Susan, M. (2007). ‘Dirty Data’ is a Business Problem, Not an IT Problem, Says Gartner. Gartner.com. Retrieved 29 December 2015, from http://www.gartner.com/newsroom/id/501733
- Unstructured Data: a big deal in big data. (2015) (1st ed., pp. 1-3). Franklin, TN. Retrieved from http://www.digitalreasoning.com/resources/Holistic-Analytics.pdf