In this research report, we present a smooth introduction to symbolic methods for knowledge discovery in databases (KDD). The KDD process is aimed at extracting from large databases information units that can be interpreted as knowledge units to be reused. This process is based on three major steps: the selection and preparation of data, the data mining operation, and finally the interpretation of the extracted units. The process may take advantage of domain knowledge embedded in domain ontologies, that may be used at every step of the KDD process. In the following, we detail three symbolic methods for KDD, i.e. lattice-based classification, frequent itemset search and association rule extraction. Then, we present three applications of the KDD process, and we end this research report with a discussion on the the main characteristics of the KDD process.