Kahve Çekirdeği Kalitesi Sıralamasında TOPSIS Yöntemi Yaklaşımı
Fatih BoyarKahve, petrolden sonra ticareti en çok yapılan ikinci emtiadır. Son yirmi yılda nitelikli kahve alanında önemli gelişmeler olmuştur. Bu gelişmelerle birlikte kahvenin tadı, çekirdeğin kusurları, üretim koşulları, sürdürülebilirlik koşulları gibi birçok konu tartışılagelmektedir. Bu çalışmada Kahve Kalitesi Enstitüsü (KKE) (CQI - Coffee Quality Institute) tarafından kabul edilen kahve kalitesi standartları ile kahveler arasında en iyi alternatifin bulunması hedeflenmiştir. KKE veri tabanında bulunan veri, ağ kazıma yöntemi ile Python 3.10.5 programı kullanılarak elde edilmiştir. Elde edilen veride aroma, tat profili, damakta kalıcılık, asitlik, gövde, tat dengesi, tutarlılık, temizlik, tatlılık nitelikleri olumlu yönde etkileyen ölçütler; 1. kategori kusur, 2. kategori kusur ve kavurma sonrası açık renk kalmış nitelikleri kahvenin kalitesini olumsuz yönde etkileyen ölçütlerdir. Alternatifler arasında en iyisinin seçimi için Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) kullanılmıştır. Ağırlıklandırma adımında hem entropi yöntemi hem de eşit ağırlıklandırma ele alınmış, olumlu ve olumsuz yönleri açıklanarak tartışılmıştır.
TOPSIS Method Approach in Ranking Coffee Bean Quality
Fatih BoyarCoffee is the second-most traded commodity after oil. Quantifying the quality of coffee is a critical topic of research. Many aspects such as the taste of coffee, bean defects, production conditions, and sustainability conditions are considered for rating the quality. In this study, the best alternative method for quality standards accepted by Coffee Quality Institute (CQI) was discussed. The web scraping method and the Python 3.10.5 program was used to obtain the data in the CQI database. Aroma, taste profile, aftertaste, acidity, body, taste balance, stability, cleanliness, sweetness qualities were believed to positively affect quality. First-category defect, second-category defect, and its light color after roasting were considered to negatively affect coffee quality. The Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) was used to select the best among the alternatives. In the weighting step, both the entropy method and equal weighting were used, and their positive and negative aspects were explained and discussed.
Coffee beans are graded on criteria such as coffee taste and aroma quality as well as visual factors such as the appearance, size, and color tone of beans. Various taste and aroma profile tests have been developed to evaluate coffee quality. Factors such as quality control factors, sustainability, working conditions during production, processing and storage of coffee beans are also considered.
Many standards have been setup in numerous organizations at the international and national level. For example, the International Coffee Organization focuses on many areas such as establishing and implementing quality standards in coffee beans, supporting and guiding coffee producers, developing sustainable coffee standards, and establishing hygiene rules.
Coffee quality institute (CQI) develops and trains cupping protocols used in coffee tasting to assess coffee quality and establish a reference for coffee tasting notes around the world. As of May 2023, more than 8,000 CQI certified grading members (CQI Graders) evaluate coffee in many countries of the world according to CQI standards.
Because of the prevalence of using CQI for grading coffee quality, this study is based on the CQI standard..
Cupping standards developed by CQI are used to evaluate coffee quality. Cupping is a standardized method of tasting coffee. CQI’s cupping standards cover the evaluation of aroma, flavor, aftertaste, acidity, body, balance, uniformity, clean cup, and sweetness.
Apart from the tasting taste grading, imperfections in coffee quality are undesirable properties that can occur in coffee beans during processing or storage and classified as category one and category two defects.
The first category of defects are primary defects that can be detected visually. Category Second defects are defects that are inconspicuous and can only be detected by taste.
Indicators of coffee quality are available in the CQI database. This database includes a unique ID number assigned to each coffee, type, origin, farm name, production altitude, harvest date, grading date, processing technique, and tasting scores.
The data of each type of coffees are kept on separate pages by using an ID number. The web scraping technique was applied to achieve rapid data collection as the data were kept on separate pages for each coffee. Network scraping was performed using Python 3.10.5. The content of the written Python program is presented in APPENDIX-1. Furthermore, web scraping files and data are available at GitHub Platform (GitHub, 2023).
The Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) method was used for ranking according to coffee grading. This method is an efficient method that is preferred in finding the ideal one in multiple preferences.
Cost attributes have a negative effect on the decision. “Aroma”, “Flavor”, “Aftertaste”, “Acidity”, “Body”, “Balance”, “Uniformity”, “Clean Cup”, “Sweetness”, “Overall” attributes, which are referred to as taste scores in the dataset, have a positive effect. However, “Category One Defects,” “Category Two Defects,” “Quakers” have a negative effect. In the obtained data set, a total of 207 observations and 13 characteristics of the mentioned coffee quality indicators are presented.
In this study, two calculations were made with the entropy weighting method, in which the attributes have equal importance.
The aforementioned methods were used, and the best choices were suggested using the TOPSIS method according to the coffee quality. A general criterion was presented with the sum of the tasting scores, but this study suggests that a calculation method that considers tasting scores separately and accounting for the error qualities could considerably improve grading.
