Art market studies have long relied on econometrics to explain the prices of works through a series of variables, such as the dimensions of the work, its medium, its date, whether or not its creator is alive, etc. However, this quantitative analysis fails to measure the qualitative opinion on the work, that can be found in exhibition or auction catalogues. As a consequence, no evidence is made between the price of an artwork and the critical comments on the latter, even if it seems obvious that such a link exists. This lightning talk aims at filling this gap by introducing Natural Language Processing, and more specifically, sentiment analysis, into econometrics (Généreux et al., 2008).The study of opinions is all the more relevant as it relates to an emerging market: the paper will thus focus on the Parisian auction sales of so-called “modern” paintings in the 19th century – as opposed to the Old Masters market. A comprehensive transcription of all the related catalogues (for more details, see Saint-Raymond, 2018) allowed to apply sentiment analysis to all the descriptions of the works sold at auction. The minutes of the sales, curated at the Archives de Paris, then provided the hammer prices, which were added to this dataset. Finally, hedonic regressions were run, including sentiment analysis (Liu, 2018) as an explanatory variable, in addition to all the traditional ones. Sentiment analysis is based on a lexical analysis taking into account the specificities of our corpus.This talk will be an opportunity to discuss our results and the relevance of this methodological encounter: how can one measure opinions on artworks and, more broadly, tastes?
If this content appears in violation of your intellectual property rights, or you see errors or omissions, please reach out to Scott B. Weingart to discuss removing or amending the materials.
Hosted at Carleton University, Université d'Ottawa (University of Ottawa)
Ottawa, Ontario, Canada
July 20, 2020 - July 25, 2020
475 works by 1078 authors indexed
Conference cancelled due to coronavirus. Online conference held at https://hcommons.org/groups/dh2020/. Data for this conference were initially prepared and cleaned by May Ning.
Conference website: https://dh2020.adho.org/
Series: ADHO (15)