A Cost-Effective Content-Based Recommendation Algorithm for Gourmet Food Preference
Jun-Cheol Suh, Jung-Been Lee, Jeong‐Dong Kim, Taek Lee · Asia-pacific Journal of Convergent Research Interchange · 2024
In this study, we propose a cost-effective gourmet food recommendation algorithm using a content-based filtering approach.This algorithm can be used to help people select foods that they enjoy and like because of their unique aroma or taste, such as whiskey or wine, for example.In our case study, we explain how the recommendation algorithm works for whiskey.First, the recommendation algorithm finds N similar data records in the database by inputting the user's quantified flavor expression data and a descriptive text and taste note about the desired flavor.During the search process, the products' price, quality rating, and similarity are used to calculate the R-value (price/benefit ratio).Next, it sorts the N neighbors according to their R-values, and finally, it puts the product with the highest R-value at the top of the recommendation list.We calculate various contextual recommendation scenarios and present the results of an efficient algorithm that suggests the most preferred product options for a given budget.To evaluate the effectiveness of our recommendation algorithm, we conducted a satisfaction survey and found that the recommendation success rate was 70%.This is a 20% performance improvement compared to the 50% success rate achieved with the collaborative filtering approach.The suggestion method will make it easier for people to find products that match their tastes and preferences at a lower cost and minimize the chance of making a wrong choice.The algorithm can be applied to many scenarios where people are looking for their favorite foods or foods with other flavors, such as wine and whiskey.