Market basket analysis, or association analysis, is only one of the many possible functions that data mining tools can perform. This kind of analysis relates to discovering of the retail sales data to enable the user to draw conclusions as to the relations between the products that are unrelated at the first sight. As it is discussed in the Beer and Diapers case, beer is most often bought along with diapers, soap and chips. Implementations of such conclusions can lead to a significant rise in sales and profits for the retailers. This can be done, for example by laying these four products in the nearly-located aisles, or by putting special offers for, let us say, soap on then packs of diapers.
The difference of general data mining techniques and the marketing basket analysis lies in the very fact that most data mining operations, as well as the statistical data always contain some degree of uncertainty. In market basket analysis, however, the relationships are not derived from these uncertain relations. Instead, this kind of analysis concentrates on deducing conclusions from the already available past data. An example of pattern discovery is the analysis of retail sales data to identify seemingly unrelated products that are often purchased together. Market basket analysis is a product of data mining.
The evidence for it is contained in the very definitions of each term. Data mining is process of efficient discovery of not obvious valuable patterns from a large collection of data, whereas market basket analysis is directed at discovering the if this then that relationships, which associate events in a database. For example the association between purchased items at a supermarket. Association rules of the market basket analysis are tightly related with general correlation rules. Correlation is a useful tool for determining if relationships exist between securities. A correlation coefficient is the result of a mathematical comparison of how closely related two variables are.
The relationship between two variables is said to be highly correlated if a movement in one variable results or takes place at the same time as a similar movement in another variable. There can be clearly seen that association rules are based on analyzing how one product is sold in relation to the other. The aim of market basket analysis is to show that whenever A occurs, B also tends to occur: A B. A useful feature of correlation analysis is the potential to predict the movement in one security when another security moves. Sometimes, there are securities that lead other securities. In other words, a change in price in one results in a later change in price of the other. A high negative correlation means that when a securities price changes, the other security or indicator or otherwise financial vehicle, will often move in the opposite direction.
The market basket analysis in not concerned with the price so much as it is concerned with predicting how event A will react to changes in the event B and visa versa. Similarly, association rule is aimed at suggesting the degree of confidence of the events happening. The higher the confidence level is, the more products correlate. Correlation, however, is not static. In other words, the correlation between two things in the markets does change over time and so even a careful understanding that what has happened in the past may not always predict what will happen in the future. The same happens with the market basket analysis, which heavily relies on the consumer behavior.
It is obviously difficult to be absolutely sure how consumers will behave, but such an analysis is still very useful for giving a good insight into a current situation in the market. The importance of using software is indisputable today. It is not only the most quick and efficient way to collect, store and analyze information, it is also the best aid in planning and decision-making. The most contemporary system, called data Mining is widely used in all business spheres. Most retailers use one of the variations of the data mining system marketing basket analysis, which helps in determining which products are correlated in sales with which. Association rules follow same patterns as the correlation principles, including the certainty coefficients, how one products sales will behave in accordance with the other product, etc.
The extended look at the market basket analysis will ensure its users of the desired outcomes of the marketing campaigns.