• Proceedings of the Korea Contents Association Conference
• /
• 2005.05a
• /
• pp.244-247
• /
• 2005

While the interest and demand of wine is increasing, the systematic support of information is becoming a mandatory requirement. In this thesis, let us suggest 'Tannin', one of the elements which is consisted in the wine taste, as one of the classification standard, and divide the classification standards of wine between the characteristics of wine and the outer environment.

• Culinary science and hospitality research
• /
• v.11 no.2
• /
• pp.32-47
• /
• 2005

As wine is increasingly becoming a lifestyle beverage among local consumers, there is a great need to understand wine consumer profiles. According to the research of this study, there is a clear evidence that four wine consumer groups exist in the Korean domestic wine market. Wine consumers are classified into four different groups: connoisseurs, aspirants, newcomers, and outsiders. This study has shown that wine with rich aroma and red wine preferred the most and young wine and white wine are the least selected in general.

• International Journal of Internet, Broadcasting and Communication
• /
• v.10 no.2
• /
• pp.25-30
• /
• 2018

This paper is about wine quality classification with multilayer perceptron using the deep neural network. Wine complexity is an issue when predicting the quality. And the deep neural network is considered when using complex dataset. Wine Producers always aim high to get the highest possible quality. They are working on how to achieve the best results with minimum cost and efforts. Deep learning is the possible solution for them. It can help them to understand the pattern and predictions. Although there have been past researchers, which shows how artificial neural network or data mining can be used with different techniques, in this paper, rather not focusing on various techniques, we evaluate how a deep learning model predicts for the quality using two different activation functions. It will help wine producers to decide, how to lead their business with deep learning. Prediction performance could change tremendously with different models and techniques used. There are many factors, which, impact the quality of the wine. Therefore, it is a good idea to use best features for prediction. However, it could also be a good idea to test this dataset without separating these features. It means we use all features so that the system can consider all the feature. In the experiment, due to the limited data set and limited features provided, it was not possible for a system to choose the effective features.

• Journal of the Korea Society of Computer and Information
• /
• v.22 no.2
• /
• pp.81-87
• /
• 2017

Nowadays wine is increasingly enjoyed by a wider range of consumers, and wine certification and quality assessment are key elements in supporting the wine industry to develop new technologies for both wine making and selling processes. There have been many attempts to construct a more methodical approach to the assessment of wines, but most of them rely on objective decision rather than subjective judgement. In this paper, we propose a data mining approach to predict human wine taste preferences that is based on easily available analytical tests at the certification step. We used sequential forward selection and decision tree for this purpose. Experiments with the wine quality dataset from the UC Irvine Machine Learning Repository demonstrate the accuracies of 76.7% and 78.7% for red and white wines respectively.

• Proceedings of the Korean Society of Near Infrared Spectroscopy Conference
• /
• 2001.06a
• /
• pp.1516-1516
• /
• 2001

During the recent years, wine analysis has played an increasing role due the health benefits of phenolic ingredients in red wine [1]. On the other hand there is the need to be able to distinguish between different wine varieties. Consumers want to know if a wine is an adulterated one or if it is based on the pure grape. Producers need to certificate their wines in order to ensure compliance with legal regulations. Up to now, the attempts to investigate the origin of wines were based on high-performance liquid chromatography (HPLC), gas chromatography (GC) and pyrolysis mass spectrometry (PMS) [l,2,3]. These methods need sample pretreatment, long analysis times and therefore lack of high sample throughput. In contradiction to these techniques using near infrared spectroscopy (NIRS), no sample pretreatment is necessary and the analysis time for one sample is only about 10 seconds. Hence, a near infrared spectroscopic method is presented that allows a fast classification of wine varieties in bottled red wines. For this, the spectra of 50 bottles of Cabernet Sauvignon, Lagrein and Sangiovese (Chianti) were recorded without any sample pretreatment over a wavelength range from 1000 to 2500 nm with a resolution of 12 cm$\^$-1/. 10 scans were used for an average spectrum. In order to yield best reproducibility, wines were thermostated at 23$^{\circ}C$ and a optical layer thickness of 3 mm was used. All recorded spectra were partitioned into a calibration and validation set (70% and 30%). Finally, a 3d scatter plot of the different investigated varieties allowed to distinguish between Cabernet Sauvignon, Lagrein and Sangiovese (Chianti). Considering the short analysis times this NRS-method will be an interesting tool for the quality control of wine verification and also for experienced sommeliers.

• Tuberculosis and Respiratory Diseases
• /
• v.81 no.4
• /
• pp.347-348
• /
• 2018

• Journal of the Korean Society of Food Culture
• /
• v.28 no.4
• /
• pp.371-385
• /
• 2013

This study aimed to classify attendants of a wine training institute according to loyalty for wine training service program and to deduce the core managing points in a wine training program by IPA (Importance-Performance Analysis). Self-administered questionnaires were collected from 192 trainees and statistical data analysis completed using SPSS ver. 18.0. As a result of clustering analysis based on trainee loyalty from both attitude and behavioral perspectives, four classification groups were identified: a "genuine" loyalty group, a "latent" loyalty group, a "mendacious" loyalty group, and a "low" loyalty group. For the genuine loyalty group, the importance of total service quality was 4.32 on average whereas the performance was measured as 4.22; thus there was little difference between importance to quality and performance. However, for the other three groups, especially the low loyalty group, there were significant wide gaps between importance to quality and performance. According to IPA, different service quality items were posted on the 'Focus here' quadrant (a domain with high service quality importance but low performance) by group, while the other three quadrants had several common items regardless of the group. Finally, the core quality managing points were different depending on the level of trainee loyalty. Therefore, it is necessary to plan and conduct a wine training program that reflects the characteristics and needs of its students, which will lead to a differentiated management strategy according to the level of loyalty.

• Proceedings of the Korean Society of Near Infrared Spectroscopy Conference
• /
• 2001.06a
• /
• pp.1257-1257
• /
• 2001

Fourier transform near infrared (FT-NIR) spectroscopy was used as a rapid method to measure the $^{o}Brix$ content and to discriminate between different must samples in terms of their fee amino nitrogen (FAN) values. FT-NIR spectroscopy was also used as a rapid method to discriminate between Chardonnay wine samples in terms of the status of the male-lactic fermentation (MLF). This was done by monitoring the conversion of malic to lactic acid and thereby determining whether MLF has started, is underway or has been completed followed by classification of the samples. Furthermore, FT-NIR spectroscopy was applied as a rapid method to discriminate between table wine samples in terms of the ethyl carbamate (EC) content. EC in wine can pose a health threat and need to be monitored by determining the EC content in relation to the regulatory limits set by the authorities. For each of the above mentioned parameters, $QUANT+^{TM}$ methods were built and calibrations derived and it was found that a very strong correlation existed in the sample set for the FT-NIR spectroscopic predictions of $^{o}Brix$ (r = 0.99, SECV = 0.306), but the correlations for the FAN (r = 0.61, SECV = 272.1), malic acid (r = 0.58, SECV = 1.06), lactic acid (r = 0.51, SECV = 1.14) and EC predictions (r = 0.47, SECV = 3.67) were not as good. Soft Independent Modeling by Class Analogy (SIMCA) diagnostics and validation was applied as a sophisticated discrimination method. The must samples could be classified in terms of their FAN values when SIMCA was applied, obtaining results with recognition rates exceeding 80%. When SIMCA diagnostics and validation were applied to determine the progress of conversion of malic to lactic acid and the EC content, again results with recognition rates exceeding 80% were obtained. The evaluation of the applicability of FT-NIR spectroscopy measurement of FAN, $^{o}Brix$ values, malic acid, lactic acid and EC content in must and wine shows considerable promise. FT-NIR spectroscopy has the potential to reduce the analytical times considerably in a range of measurements commonly used during the wine making process. Where conventional FT-NIR calibrations are not effective, SIMCA methods can be used as a discriminative method for rapid classification of samples. SIMCA can replace expensive, time-consuming, quantitative analytical methods, if not completely, at least to some extent, because in many processes it is only needed to know whether a specific cut off point has been reach or not or whether a sample belongs to a certain class or not.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.38 no.1
• /
• pp.139-152
• /
• 2013

The purpose of this study is to claim the validity of tuning the architecture of neural network models for multi-class classification. A neural network model for multi-class classification is basically constructed by building a series of neural network models for binary classification. Building a neural network model, we are required to set the values of parameters such as number of hidden nodes and weight decay parameter in advance, which draws special attention as the performance of the model can be quite different by the values of the parameters. For better performance of the model, it is absolutely necessary to have a prior process of tuning the parameters every time the neural network model is built. Nonetheless, previous studies have not mentioned the necessity of the tuning process or proved its validity. In this study, we claim that we should tune the parameters every time we build the neural network model for multi-class classification. Through empirical analysis using wine data, we show that the performance of the model with the tuned parameters is superior to those of untuned models.

• Culinary science and hospitality research
• /
• v.23 no.6
• /
• pp.143-152
• /
• 2017

This paper examined seemingly well-working compromises in identity construction, questioning whether the compromises could function only nominally in practice. The literature has paid attention to the conflicts which end up functionally sense-making, through either unilaterally enforced or mutually assimilated compromises. In contrast, this paper's analysis of Super Tuscan wines under the Italian government's quality regulation illustrated that the compromises between wineries and classification systems do not work well and make the classification systems meaningless in the end. This study thus argued that compromises in identity construction do not always result in functionally sense-making outcomes: they could be only nominal. This study suggested that idiosyncratic institutional contexts, such as weak organizational legacy, affect the results of identity construction in functional terms. At last, the theoretical and practical implications both in organization and management of this study were well discussed.