This paper presents an application scenario of a rule-based recommender model (LIME) to a PSLC cognitive tutor analytics database. LIME is a recommendation model which provides a formative support to students, tutors and lecturers, thanks to a visual interaction between users and the elearning system. On the other side, cognitive computer tutors, kindly available from the Pittsburg Science Learning Centre (PSLC) and the Carnegie Mellon University (CMU) learning analytics repository, provide intelligent and visual interactive activities that help middle-school students improve their daily academic skills. They also offer on-demand step-by-step feedback during problem solving and report on student progress for teachers and students. Every single learner interaction with a tutor is logged and is freely available within specific databases expressed in the PSLC Logging Format (also known as Tutor Message format). This format is well defined in an XML Schema Definition (XSD file) and a Document Type Definition (DTD file). We begin our text by presenting all these technologies (LIME model, tutors and the PSLC Logging Format). We then forge some example configurations of our recommender model, and apply them to a database of over 1250 students, 2250 interaction hours and almost 1 Gb of logged data. The result of this process is a potential recommendation to each of these students after the use period of the interactive tutor software. Finally, we present the results of this practical implementation, and discuss the validity or our model in the related tutoring environments.