Concepts such as typicality are difficult to demonstrate using the limited set of samples that can be subjected to sensory evaluation. This is due both to the complexity of the concept and to the limitations of traditional sensory evaluation (number of samples per session, panel fatigue, the need for multiple sessions and methods, etc.). On the other hand, there is a large amount of data already available, accumulated through many years of consistent evaluation. These data are held in repositories (such as Platter’s Wine Guide in the case of South Africa Wine, wineonaplatter.com) and in technical notes provided by the producers.There are many unknowns regarding the distinguishing features of a commercial Old Vine (OV) Chenin Blanc wine and its comparison to a Young Vine (YV) wine. There is little work done on it and the work has limitations regarding the methodology and number of samples included (Crous, 2016; Mafata, Brand, Panzeri, et al., 2020). Platter’s data contains descriptors for wines produced in South Africa, as well as a quality rating. The producers put technical sheets together – while the expert tasters generate Platter’s data for the same wines.Similar to work done on the general characteristics of South African Chenin Blanc wine (Valente, Bauer, Venter, et al., 2018), the goal of the study is to find the unique features associated with the ‘old vine Chenin Blanc character’ using available data from expert tastings and technical notes. During the initial step, Platter’s data and technical notes are mined for attributes of Chenin Blanc wines (as both sources indicate whether the wines belong to the Old Vine category). The automated process is done using the data gathering and analysis tool developed by the research team. A combined data set from all data sources is also created.During the analysis step, Agglomerative Hierarchical Clustering (AHC), Multiple Correspondence Analysis (MCA), Fuzzy K-Means clustering (FKM), and Formal Concept Lattice (FCL) are employed to explore the attribute and product space. Clustering algorithms are applied to the data (separate and fused sets) to identify markers (features) for the Old Vine character. As Platter’s data also includes product ratings, the possible correlation of Old Vines vs. Young Vines regarding the perceived quality can also be tested. In addition to finding sensory attributes associated exclusively with Old Vine Chenin Blanc (the typicality issue), the novelty of the work also resides with the creation and development of a new application for the automated data gathering and analysis tool, whose effectiveness and robustness will be tested in the real case scenario.
Authors: Kruger Markus1, Brand J.1, Watson B.2, Mafata M.1 and Buica A.1
1Department of Information Science, Stellenbosch University, South Africa; South African Grape and Wine Research Institute, Department of Viticulture and Oenology, Stellenbosch University, South Africa
2 Department of Information Science, Stellenbosch University, South Africa
*corresponding author: 11368446@sun.ac.za
References:
Crous, R. 2016. The sensory characterisation of old-vine Chenin blanc wine: an exploratory study of the dimensions of quality. Stellenbosch University.
Mafata, M., Brand, J., Panzeri, V. & Buica, A. 2020. Investigating the Concept of South African Old Vine Chenin Blanc. South African Journal of Enology and Viticulture. 14(2):168–182.
Valente, C.C., Bauer, F.F., Venter, F., Watson, B. & Nieuwoudt, H.H. 2018. Modelling the sensory space of varietal wines: Mining of large, unstructured text data and visualisation of style patterns. Scientific Reports. 8(1).
Keywords: Chenin Blanc, Old Vine, Automation, Multi-source data gathering
