Developing an Application to Score Unreviewed Wines; How the Delicacy of Taste Enables Expert Wine Reviews to Function as a Placebo with Respect to the Consumer Purchasing and Tasting Wine Experience 324 views

The global wine industry has seen massive growth in recent years and with that growth there have been many novice consumers. As being a novice implies, new consumers need help to pick out good bottles of wine and the most prominently used aid is expert wine reviews, generally given as a score out of 100 indicating the quality of the wine. Both of my projects focused on the prominence of the use of expert wine reviews by consumers. For the technical project, I created a machine learning algorithm create a score for a bottle of wine that does not yet have an expert created score using only information on the label of the bottle. My STS research focused on the role of expert wine reviews in the wine industry as a placebo to increase consumer tasting enjoyment and to reduce consumer perceived purchasing risk. The large amount of online data available on expert wine scores makes it possible to use the data to train an accurate machine learning model to predict a wine score. The technical portion of my thesis produced a prediction model that synthesized a database of over 150,000 expert reviews to be able to predict a bottle’s score based on label information including country of origin, region, price, province, vintage, and variety. The model can correctly predict a bottles expert rating given the previous features to within ± 2.007 of the given experts score. This accuracy is good enough for a consumer to be able to be confident about a bottle’s quality, even if it is not accompanied by an expert created review. The delicacy of taste helps explain why experts are better able to assess the quality of a wine over a novice. In my STS research I used the framework of the delicacy of taste to evaluate the role of expert reviews in the wine industry. In my research I found the expert reviews act as a placebo effect in decreasing the risk consumers feel when purchasing wine as well as increasing the perceived quality of the wine when tasting. This result gives insight into when expert wine reviews are relevant in the wine industry and provides a new way to look at their relevancy to the industry. The work done in both the technical and STS portions of my thesis worked together to give me a better understanding of how expert wine reviews function in the wine industry. Without doing both portions I would not have had a full synthesis of how expert reviews are both created and deployed in the industry. My project exemplifies the themes put forth in the STS curriculum as it shows the gravity of effect that expert reviews can have on consumers. The technical power to create reviews can have serious consequences as I show how impactful a good (or bad) review can be on consumer experience.

BS (Bachelor of Science)

Expert Wine Reviews; Wine Rating Placebo; Wine Rating Creation; Wine Rating Machine Learning

School of Engineering and Applied Science Bachelor of Science in Computer Science Technical Advisor: Charles Reiss STS Advisor: Kathryn Neeley Technical Team Members: Jared Hood

English

All rights reserved (no additional license for public reuse)

2021-05-15