This work deals with the issue of understanding a user's behaviour as this is expressed via a gamified application. The notion of ontologies and the association of concepts in relevance to decisions that have to be made is used. The current work introduces a new process-based approach, based on collected large log files and associations of underlying decisions based on them. Both of them deal with work in extracting information for intelligent use, the main difference being that the first discovers but stops on a concept relation basis, while the other based on processes, as knowledge transactions, further to the associations on a 1:1 level maybe applied on a multi associative model. The objective of the current work is to introduce the methodology into gamified environments (such as but not limited to) games, for semi-automated understanding of user behaviour and furthermore, prediction and in instances, guidance via optimal paths of decision making activities, that are useful in gamified applications in various areas like the education. Both the initial ontology based, and the extension work on it, are based on mining association rules, in one instance treated as knowledge nodes (concepts) and in the second as underlying knowledge processes, based on big log files. This may be applicable to online games, that generate big log files of user selections, that are available for study and examination for extracting user behavioural patterns. As a result, maximum length of sequential patterns and items in them, are discovered in an algorithmic based methodological approach, providing in this way a set of guidelines for designing gamified applications.