Data Mining and Recommendation Engines

Data mining is the process of analyzing large amounts of seemingly disparate data that have been collected through various means to reveal new information through patterns or other methods. Data mining is used in many fields, such as business, science, engineering, medicine, and game design to inform a wide array of processes such as research, business intelligence, and marketing. Data mining is commonly performed on data that are collected about individuals as they interact with online or other networked computer systems. These data may be intentionally provided by the person through profiles and surveys, for example, or may be collected as metadata through automated processes, such as collecting a user’s IP address, browser type, interactions within a system, or even information about other websites a person accesses. The large-scale data collection about individuals is commonly referred to as big data and is widely employed by business and other industries to determine the characteristics, needs, and desires of individuals. While there are understandable and growing ethical concerns about big data in regard to individuals’ privacy and possible misuse, the functional applications of data mining reach into nearly every aspect of modern society, including education. This entry discusses how data mining and recommendation engines work, how they can be used in education, and some of their implications for education.

The principles of data mining are rooted in several related fields such as artificial intelligence and machine learning. The patterns that are revealed through data mining allow specific functions to be performed based on those patterns. Essentially, through data mining processes, a system is able to learn about an individual through the data and metadata that are available about the person. A common application of data mining is known as a recommendation engine, also referred to as a recommender system. A recommendation engine is a type of information filtering system that uses data mining algorithms to determine targeted suggestions to an individual for additional information or resources that may serve the immediate or future needs of an individual. An example of a recommendation engine is when a person shops at an online store, the shopper is presented with additional recommended items that the individual may decide to purchase. These recommendations may be based on any number of variables, such as other items viewed or purchased by the person, demographic characteristics about the person (e.g., gender, age, location), and items purchased by others who also viewed or purchased the same items as this individual, just to name a few examples. Another example of a recommendation engine in action is when a person searches for information about a product online and then begins to frequently see advertisements for that item while visiting other websites.

Recommendation engines employ complex algorithms to produce recommendations based on the patterns revealed through the myriad data available through data mining. This presentation of recommended resources can be a tremendous benefit to users who would otherwise be overwhelmed by the endless amounts of information available to them. Instead of expecting a person to seek out useful and relevant resources, these resources are brought to the person, including ones that the person would not otherwise be aware of and, therefore, would not have even known to search for. The end result can be increased efficiency and appeal of using the system for the user and also increased acquisition and utilization of resources or [Page 174]services being provided, which benefits both resource users and providers. Because of these benefits, recommendation engines are nearly pervasive among online systems such as search engines, e-commerce; social media; music, movie, and book sites; and online research databases, just to name a few.

...