Companies have been generating and accumulating a lot of data, and the
current transformation of society requires that these organizations learn to constantly
deal with data. Data exists without information, but information only exists through
data, and the type of relationship that an organization establishes with this information
is fundamental to understanding the behavior of its customers. Through Data Mining it
is possible to extract information from the modeling and analysis of data, generating
insights from the day-to-day business for a better decision by the managers of an
organization. Based on this, this work seeks to find in the literature an algorithm
capable of performing the task of Market Basket Analysis, aiming at the generation of
association rules from a database of sales transactions in a convenience store.
Subsequently, these generated rules are related to the days of the week, providing
useful information on the daily behavior of customers at the point of sale. In order for
the data mining stages to be achieved, the work follows the CRISP-DM methodology,
exposing everything from understanding the business to evaluating the results. With
the application of the steps of the methodology and the model, it was possible to
discover which products tend to be sold together in certain periods of the week, in
which these results bring insights to support the decision-making of those responsible
for the business in the search for differentiation in front competition. The study proves
that the methodology and the tool used can be important means to, in the current
competitive market, generate information of great importance to organizations.
