Understanding South African Chenin Blanc wine by using data mining techniques applied to published sensory data

Valente, Carlo Cesar (2016-03)

Thesis (MSc)--Stellenbosch University, 2016.

Thesis

ENGLISH ABSTRACT: South African Chenin Blanc is the most planted grape cultivar in South Africa (SA) and is known for its versatility in wine sensory profiles. However, according to the South African wine industry, consumers are confused as to the different styles that make up Chenin Blanc wine. Currently, there are six different style classifications for South African Chenin Blanc wine that was proposed as a guideline by the Chenin Blanc Association (CBA). Previous research conducted at the University of Stellenbosch was aimed at evaluating these style classifications. Previous results showed that, when using a small sample set of commercial Chenin Blanc, only two clear style categories could be identified – Fresh and Fruity and Rich and Ripe Wooded. However, there was a need to re-evaluate these style classifications using a larger sample set. The current study proposed a novel approach of using reputable and published text data on a Chenin Blanc wines as a dataset as opposed to acquiring a limited amount of Chenin Blanc wines for sensory evaluation. The John Platter Wine Guide is South Africa’s largest wine annual and provides a sensory description of each wine. Text data on Chenin Blanc was used over an investigation period of seven years (2007-2014) as it was felt that this could give a clear indication of the sensory properties within the cultivar. It would also provide us with an extremely large dataset to evaluate the current style classification of Chenin Blanc and this was one of the main aims of the study. Computed text data analysis on 2746 Chenin Blanc wines showed that there were only two clear style categories identified – Fresh And Fruity and a Wooded style, which was conclusive with previous research conducted. Additionally, for the first time the entire sensory spectrum of South African Chenin Blanc was characterised and identifiable “cues”, as mentioned in the wine industry, was established. For the purpose of validation, a second white wine cultivar, Sauvignon Blanc, was used to evaluate whether text data analysis on Chenin Blanc was conducted correctly. Furthermore, the sensorial properties that were retrieved from the dataset for the characterization of the full sensory spectrum were later used for the re-evaluation and validation of the current Chenin Blanc Aroma Wheel. The Chenin Blanc Aroma Wheel was developed in 2005 and since then has never been re-evaluated. There has always been a need to re-evaluate the Aroma Wheel, but there has never been a large enough dataset to do so. This then served as a novel approach as it was the first time that such a task would be attempted. A frequency count of all the words used in the Platter Guide was used in conjunction with sensory attributes that currently made up the aroma wheel. An exclusion/inclusion attribute criterion of five citations was set. Attributes that were currently on the Aroma Wheel that were not cited at least five times were excluded and those that were currently not on the Aroma Wheel and were cited five times or more were included. Additional categories and sub-categories were included to better accommodate attributes that were previously presented in the wrong category. Figure 1 illustrates the current Chenin Blanc Aroma Wheel while figure 2 illustrates the validated Aroma Wheel as proposed in this study. Even though the two Aroma Wheels share similarities, it is evident that, for South African Chenin Blanc, the current aroma wheel is not an accurate representative of the white wine cultivar.

AFRIKAANSE OPSOMMING: Suid-Afrikaanse Chenin Blanc is die mees volop verboude druifkultivar in Suid Afrika (SA) en is bekend vir veelsydigheid in terme van wyn sensoriese profiele. Nietemin is daar volgens die Suid- Afrikaanse wynindustrie ‘n verwarring onder verbruikers rakende die verskillende Chenin Blanc wynstyle. Tans is daar ses verskillende stylklassifikasies vir Suid Afrikaanse Chenin Blanc wyne soos voorgestel in die Chenin Blanc Assosiasie (CBA) se riglyne. Vorige navorsing deur die Universiteit van Stellenbosch was gemik daarop om hierdie stylklasse te evalueer. Vorige resultate het getoon dat daar binne ‘n klein stel monsters van kommersieële Chenin Blanc wyne slegs twee duidelike stylklasse geindentifiseer kon word: Vars en Vrugtig sowel as Ryk en Ryp Gehout. Dit was noodsaaklik om hierdie beperking in stylklasse weer te evalueer binne die konteks van ‘n groter stel monsters. Die huidige studie stel ‘n nuwe aanslag voor waartydens betroubare en gepubliseerde teksdata oor Chenin Blanc wyn as ‘n datastel gebruik word in plaas daarvan om ‘n beperkte hoeveelheid Chenin Blanc wyne sensories te evalueer. Die John Platter Wyngids is Suid Afrika se grootste jaarlikse wyngids wat ‘n sensoriese beskrywing van elke wyn verskaf. Teksdata oor Chenin Blanc was gebruik oor ‘n ondersoekperiode van sewe jaar (2007-2014) ten einde ‘n duidelike aanduiding van die sensoriese eienskappe wat in Chenin Blanc voorkom te bekom. Berekende teks data-analise op 2746 Chenin Blanc wyne het getoon dat daar wel slegs twee duidelike stylklasse geidentifiseer kon word: Vars en Vrugtig sowel as Gehout. Hierdie bevinding is ooreenstemming met vorige bevindinge. Benewens is die hele sensoriese spektrum van Suid- Afrikaanse Chenin Blanc wyne gekarakteriseer en die identifiseerbare geur nuanses soos gebruik deur die wynindustrie is bevestig. Ten einde die resultate te bekragtig is ‘n tweede witwynkultivar, Sauvignon Blanc, deur dieselfde teks data-analise geanaliseer. Verder is die sensoriese eienskappe wat verkry is vanuit die datastel vir die karakterisering van die volle sensoriese spektrum weer gebruik om die huidige Chenin Blanc Aroma Wiel te herevalueer en te bekragtig. Die Chenin Blanc Aroma Wiel was sedert dit in 2005 ontwikkel is nog nie weer hersien nie. Ten spyte van die behoefte om die Aroma Wiel te herevaleer kon die taak, weens ‘n gebrek aan ‘n datastel van voldoende grootte, tot en met nou nog nie verrig word nie. Hierdie studie is die eerste maal dat so ‘n taak op hierdie wyse aangepak word. ‘n Frekwensietelling van al die woorde wat in die John Platter Gids gebruik is is vergelyk met die sensoriese eienskappe wat teenwoordig is op die huidige Chenin Blanc Aroma Wiel. ‘n Uitsluitings/Insluitings eienskapskriteria van ten minste vyf aanhalings was vasgestel. Gevolglik is eienskappe wat nie op die bestaande Aroma Wiel voorkom nie, maar ten minste vyf keer voorkom in die Platter Gids-data, ingesluit in die hersiende Aroma Wiel. Addisionele kategorieë en subkategorieë is ingesluit in die hersiende Aroma Wiel om eienskappe wat voorheen in verkeerde kategorieë geplaas is beter te akkomodeer. Figuur 1 wys die huidige Chenin Blanc Aroma Wiel terwyl Figuur 2 die hersiende Aroma Wiel soos voorgestel in hierdie studie illustreer. Hoewel daar ooreenkomste is tussen die twee Aroma Wiele is dit duidelik dat die huidige Aroma Wiel nie ‘n akkurate weerspieëling is van Suid Afrikaanse Chenin Blanc as ‘n druifkultivar is nie.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/98866
This item appears in the following collections: