Recommender Systems; Predictive Data Analysis
“Recommender” or “Recommendation” systems refers to a type of information filtering that aims at predicting a users choice or outcome in a decision. These systems use various classification algorithms and techniques; such as decision trees for precise item choices or clustering for similarities it items to be chosen (which will be discussed shortly). Recommender systems are typically split into content-based or collaborative filtering systems:
- Content based focuses upon “properties of the items recommended”[1]. An example being ‘recommending’ content to users such as Netflix surfing or Amazon.com shopping.
- Collaborative filtering “recommend[s] items based on similarity measures between users and/or items”[1]. This type of recommender system is closely related to clustering.
Predictive Analytics stems from recommender systems which is “used to make predictions about unknown future events”[2]. This type of data analytics is anticipatory; relying on data mining and statistical techniques in order to develop these underlying insights. Predictive analytics allows for organizations to be much more forward-thinking and proactive in their business decision-making.
Recommendation system models relate closely to the topics of Decision Trees, Clustering, and Association Rule Mining that we have learned in Data Analytics (MIS2502). Multi-tiered decision tree analysis is used when recommending the best choice for an end user. Clustering is used in collaborative filtering methods in order to best group similar items with end users. Finally, association rule mining–as well as data mining in general–is used to best predict an end users ultimate choice.
As part of my job as a prospect research analyst, it is my job to determine who are the best candidates for outreach to support the school of medicine. If a recommender system algorithm were implemented, I would be able to more accurately determine which patients and doctors are more willing to contribute than others through logical decision-making analytics.
References
[1]Recommendation Systems, Chapter 9 – Stanford InfoLab n.d.. April 28 2018 http://infolab.stanford.edu/~ullman/mmds/ch9.pdf
[2]Imanuel. “What Is Predictive Analytics ?” Predictive Analytics Today, Bigtexts.com, 12 Apr. 2018, www.predictiveanalyticstoday.com/what-is-predictive-analytics/
