Home // HUSO 2017, The Third International Conference on Human and Social Analytics // View article

In the Eye of the Beer-Holder. Lexical Descriptors of Aroma and Taste Sensations in Beer Reviews

Authors:
Els Lefever
Liesbeth Allein
Gilles Jacobs

Keywords: Corpus linguistics; Natural language processing; classification; beer reviews

Abstract:
Western languages do not dispose of a well elaborated vocabulary for describing smell and flavour sensations. We investigate whether beer experts share a common vocabulary to describe beer properties. We collected an English text corpus of beer reviews and analyzed the lexical descriptors used by beer-tasting experts to describe aromas and flavours. The informativeness of beer reviews was investigated by running a machine learning experiment for predicting the colour of a beer based on the review text. This preliminary experiment shows promising results, with average accuracy figures of about 60% for automatic beer colour prediction. Our experimental results show that beer experts share a common vocabulary to describe beer characteristics in a consistent way, allowing to automatically predict beer properties based on the review text.

Pages: 25 to 29

Copyright: Copyright (c) IARIA, 2017

Publication date: July 23, 2017

Published in: conference

ISSN: 2519-8351

ISBN: 978-1-61208-578-4

Location: Nice, France

Dates: from July 23, 2017 to July 27, 2017