A Data Mining System for Fragrance Companies : Principles, Applications and benefits
Generally, data mining (sometimes called data or knowledge discovery) is the process of analyzing data from different perspectives and summarizing it into useful information.
Data mining is a technique. It concentrates on identifying patterns and relationships within large databases through the use of advanced statistical methods .
Data mining software is one of a number of analytical tools for analyzing data. It allows users to analyze data from many different dimensions or angles, categorize it, and summarize the relationships identified.
Data mining system permits to transform huge data within databases to " information and to converts this information into knowledge, in finding the relationships and patterns among dozens of fields in large relational databases.
What can data mining do for fragrance companies?
Data Mining system enables the fragrance and perfumery companies to determine relationships among " internal " structures of perfume formulas, in order to find the common olfactory notes among different aromatic raw materials used in many perfume / fragrance compositions or used in a specific perfume. Therefore, it will be possible to establish from the stored data, some relationships among different components of perfumes, and designing an aromatic benchmark or aromatic models .
Regarding aromatic raw materials and perfumes, generally speaking, four relationships can be used to transform data stocked in the databases of a fragrance company into information and industrial and technical aromatic knowledge .
These four relationships are:
1- Classes
2- Clusters
3- Associations and knowledge discovery
4- Modelling Process
Classes
Stored data concerning the process of perfumes composition , can be used to locate data in predetermined groups of aromatic raw materials.
Or to compare many pyramidal configurations of the perfumes . So from this comparison , the master perfumer can compare the configuration of the components of top notes, middle notes and base-notes, either for many perfumes or for several laboratory tests, in order to evaluate the results of the behaviours of aromatic components in perfume mixture.
Such as comparison brings a valuable information about the cost / benefits of each calibration in the structure of perfume components
Clusters
Data items can be grouped according to logical relationships or according to some criteria of classification. For fragrance companies, these criteria are , for example, the perfume family , odours family, or by the characteristics or the specifications of aromatic raw materials.
Associations of knowledge discovery
Fragrance data stocked within different data bases and information systems of a fragrance company can be mined to identify associations existing in many segments of activity, for example:
Associations between two scents or fragrance notes ,
Associations between the structure of a raw material and the structure of a perfume under composition.
The associations that can be found among the components of a perfume .
The possible synergy that can be existed between two aromatic raw materials, in order to determine the optimum of their relative importance , in case of process of calibration quantitative and qualitative .
The Modelling Process
In order to construct aromatic knowledge models, aromatic data can be mined within the database and information systems,
The traditional structure of a perfume is called olfactory pyramid and is divided into three groups that correspond to the different scents exhale scents over time:
1- The top note: the most volatile notes that feels just after spraying perfume. It is a "fresh and green note" that can last up to 2 hours.
2- The heart note: it develops for several hours and is the characteristic smell of the perfume.
3- The base note: it evaporates slowly, sometimes after several days. Its function is to fix the perfume, to make it last longer.
Every perfume has a distinct model. Its formula is based on technical coefficients of these components.
The formula of a perfume represents a knowledge scheme of the components represented by different aromatic raw materials and fixators.
The technique of data mining applications gives the possibility to discover the relationships that exists among components of several formulas.
The following are benefits of Data Mining System for fragrance companies:
I - Data mining helps to analyze and modelling data, in order to identify common patterns that exist in the composition of perfumes.
II- The patterns discovered permit to learn new technical things and evaluate the performance of the fragrances manufacturing system. Each technical discovered knowledge gives the possibility to improve the process of creativity in the field of composition and by consequence the tests in laboratory weight the preferable and the best aromatic components combinations.
III-The applications of Data Mining allow to realize the control and comparison of technical coefficients, represented by the proportion of each aromatic component contained in the formulas of a determined perfume.
So, the discovered knowledge could increase the efficiency of the production system.
IV- The Senior perfumer can construct a " model " for a perfume before starting the composition activity. This model represents the main olfactory orientations to design the perfume formula.
V- The statistical inference methods applied through data mining process gives the opportunity to construct " a priori " an olfactory model for a determined perfume, and after the end of composition process, it will be possible to identify any successful models in previous experiences.
To sum it all, the main benefit of data mining lies within the possibility of modelling the aromatic knowledge and their accumulation and performance.
The modelling of the components of a perfume facilitates in a systematic way both the design of perfume formulas and the process of perfume composition.
