Database Management Research Proposal of Wine Quality

...

5.)Also another paper“Predicting the Quality and Prices of Bordeaux Wine” also researched on price prediction of Bordeuax wine specifically. The article shows that the variability in the quality and prices of Bordeaux vintages is predicted by the weather that created the grapes. The price equation provides a measure of the real rate of return to holding wines (about 2–3% per annum) and implies far greater variability in the early wine prices than is observed. This study would shed us light on assessing market inefficiency, the effect of climate change on the wine industry and the role of expert opinion in determining wine prices.

6.)On molecular basis, there are also studies showing that Grade A wine quality is defined by certain characteristics. This paper called “The Molecular Basis for Wine Grape Quality-A Volatile Subject” showed that volatile organic compounds are important flavor components of finished wines. By studying how berry flavor components are determined by the interplay of vine genotypes, the environment, and cultivation practices at the molecular level, scientists will develop advanced tools and knowledge that will aid viticulturalists in consistently producing balanced, flavorful berries for wine production.

Wine qualities is also affected by the techonoloy Wine aging is an important process to produce high-quality wines. This paper reviews recent developments of the wine aging technologies. The impacts of operational parameters of each technology on wine quality during aging are analyzed, and comparisons among these aging technologies are made. In addition, several strategies to produce high-quality wines in a short aging period are also proposed.

---------------------------------------------------------------

Scribbles:

Data Source:

Related Articles:

Cortez, Paulo, et al. "Modeling wine preferences by data mining from physicochemical properties." Decision Support Systems 47.4 (2009): 547-553.

http://www.sciencedirect.com/science/article/pii/S0167923609001377

Reynolds, A. G., R. M. Pool, and L. R. Mattick. "Influence of cluster exposure on fruit composition and wine quality of Seyval blanc grapes."VITIS-Journal of Grapevine Research 25.2 (2015): 85.

http://pub.jki.bund.de/index.php/VITIS/article/view/5938

3. http://www.wseas.us/e-library/conferences/2009/georgia/CCI/CCI08.pdf (Aiena)

Twisha ( 4 literature Survey papers abstract)

1.)Wine Vinification prediction using Data Mining tools

The vinification procedure is one essential phases of the wines production that can impact the accomplishment of wines quality. Based in a chemical samples this appraisal is customarily acknowledged by wine testers that break down some subjective parameters, for example, color, foam, flavor and savour. This kind of investigation is critical for the production of wine and for its effective marketing. The utilization of Data Mining procedures in this field has an extraordinary significance in uncovering the significance of the various substance parameters required during the time spent wine generation. This paper shows the Decision Trees, Artificial Neural Networks and Linear Regression as Data Mining systems to accomplish the destinations of arrangement and relapse to make models to foresee the organoleptic parameters from the chemical parameters of the vinification procedure. The investigations were arranged utilizing the new Microsoft's SQL Server 2008 Business Intelligence Development and an open-source Data Mining tool WEKA. Good results were accomplished with exactnesses somewhere around 86% and 99% obtained for all models.

2.) Classification-based Data Mining Approach for Quality Control in Wine Production

Modeling the complex human taste is a vital concentration in wine enterprises. The principle motivation behind this study was to foresee wine quality in view of physicochemical information. This study was likewise directed to recognize exception or abnormality in test wine set to identify adulteration in wine. In this venture, two substantial separate datasets are utilized, which contains 1, 599 occurrences for red wine and 4, 989 occasions for white wine with 11 characteristics of physiochemical information, for example, alcohol, Ph and sulfates. Two grouping calculations, Decision tree and Naïve Bayes are connected on the dataset and the execution of these two calculations is looked at. Results demonstrated that Decision tree (ID3) beat Naïve Bayesian methods especially in red wine, which is the most well-known sort. The study also demonstrated that two properties, liquor and volatile acidity contribute profoundly to wine quality. White wine is additionally more touchy to changes in physicochemistry rather than red wine, henceforth more amount of taking care is fundamental.This research concludes that classification approach will give rooms for corrective measure to be taken in effort to increase the quality of wine during production.

3.) Predicting quality of wine based on chemical attributes

Ordinal information structure is hard to influence upon because of the way that it has both the properties of regression and classification. In this paper, a straightforward new strategy that influences upon ordinal information structure of the data: Additive Logistic Regression (ALR). This technique is utilized to anticipate the nature of Portuguese white wine in light of the chemical properties of the wine. We contrast ALR with non-parametric techniques. The outcomes demonstrated that the ALR strategy, though basic, out-plays out the non-parametric methods.as k-closest neighbor can be profoundly flexible. For this situation, it is believed that additive logistic regression improves at utilizing the ordinal structure of the data and henceforth delivers better results.

With respect to weighted linear regression, we have to note that weighted linear regression performs well when the number of predictors is less. On account of 10 factors, the indicator space might be excessively scanty, making it impossible to create great results. (This can likewise be clarified by the curse of dimensionality).

Notwithstanding better precision in expectation, we get to introduce an extra advantage of utilizing additive logistic regression: its great interpretability. As we go through the ranks, we see that