When we consider Big Data, we can see the value it provides to companies both private and public. Data is at the critical intersection of many daily interactions, and every day decisions are impacted by analysis gathered from data. The last 20 years has seen the rapid growth of e-commerce, a system that relies heavily on the collection of data.
As consumers, we are oversaturated with products and services to purchase, and our impulses to do that are satisfied when those very products and services have been catered to our preferences. Without the traditional sales associate to fill the role, recommender system technologies have become an important and integral part of the successful e-commerce business model. The ability to track the preferences and purchase patterns of users allows companies to increase customer satisfaction – and in turn boost profitability – by suggesting products with similar characteristics for comparison, recommending additional products that work well as a combination, as well as bringing attention to lesser known products that may cater to the user’s taste. While recommender systems are most commonly seen in the ‘for-profit’ business enterprise (such as Amazon, Netflix, or Spotify), there are also a number of broader, non-profit uses (such as Goodreads).
In this project, we used data scraped from Food.com and registered with Kaggle. The data consisted of recipes with unique user ids and reviews. Using this data, our goal was to provide recipe recommendations based on ingredient preferences. Presented in the form of a Shiny App, we built a content-based filtering system, in which the user is prompted to select their ingredient preferences and the recommender engine outputs closely associated recipes.