Predicting and Visualizing Wine Characteristics Through Analysis of Tasting Notes from Viewpoints

When describing complex characteristics of a specific genre, specialist expressions are often used. This can become quite a problematic situation for an inexperienced person, as expressions not used in everyday language are difficult to understand. This is particularly apparent when trying to describe wines, known as winespeak, as a range of specialist expressions are used in a subjective manner. In this paper, we propose that the descriptions of wines can be analyzed from various points of view to automatically predict and visualize the sensory sentiment characteristics described within the expressions as a radar chart. This would enable those not knowledgeable in winespeak to visualize and compare the complex descriptions often found in expert tasting notes.