Unit:
Department of Music StudiesLibrary of the School of Philosophy
Dissertation committee:
- Αναστασία Γεωργάκη, Αναπληρώτρια Καθηγήτρια, Φιλοσοφική Σχολή, Τμήμα Μουσικών Σπουδών, ΕΚΠΑ
- Χριστίνα Αναγνωστοπούλου, Αναπληρώτρια Καθηγήτρια, Φιλοσοφική Σχολή, Τμήμα Μουσικών Σπουδών, ΕΚΠΑ
- Σμαράγδα Χρυσοστόμου, Καθηγήτρια, Φιλοσοφική Σχολή, Τμήμα Μουσικών Σπουδών, ΕΚΠΑ
- Ίρμγκαρντ Lerch – Καλαβρυτινού, Καθηγήτρια, Φιλοσοφική Σχολή, Τμήμα Μουσικών Σπουδών, ΕΚΠΑ
- Αρετή Ανδρεοπούλου, Επίκουρη Καθηγήτρια, Φιλοσοφική Σχολή, Τμήμα Μουσικών Σπουδών, ΕΚΠΑ
- Σταματόπουλος Παναγιώτης , Επίκουρος Καθηγητής, Σχολής Θετικών Επιστημών, Τμήμα Πληροφορικής και Τηλεπικοινωνιών, ΕΚΠΑ
- Καμπουρόπουλος Βασίλειος-Αιμίλιος, Αναπληρωτής καθηγητής, Σχολή Καλών Τεχνών, Τμήμα Μουσικών Σπουδών, ΑΠΘ
Original Title:
Computational Music Analysis of Children's Improvisations: A Data Mining Approach
Translated title:
Computational Music Analysis of Children's Improvisations: A Data Mining Approach
Summary:
Music improvisation is lately gaining considerable attention, as a skill that should be
cultivated and promoted through the educational music process. Once a dexterity
rather neglected, it is now recognised as a skill of significant importance in the
development of musical abilities. Hence, children have been encouraged to
improvise during their musical classes and new teaching techniques and tools have
emerged, advocating the whole improvisation process and aiding both parties,
students and tutors, throughout the training course.
These techniques have been developed towards the teaching process as well as the
assessment of the progress of the children, and provide qualitative and quantitative
measures in order to evaluate and assist children’s improvisation efforts. With the
introduction of informational technology, such tools have become sophisticated and
automate the whole process; they provide at the same time the means for further
analysis of the improvisations, by collecting the recordings, analysing the data and
pinpointing at various interesting factors for further analysis.
The research reported in this thesis, has been conducted within the EU MIROR FP7
project. In the course of the project, a number of psychological experiments were
performed, including a number of improvisations of children, between 4 and 8 years
old. The improvisations were performed on a MIDI keyboard and the resulting data
collected and analysed in a number of ways. The aim was on the one hand to identify
significant patterns in the music produced and on the other to come up with a model
of assessing the creativity embedded in those improvisations.
The results are explored towards a three-fold goal: (i) the identification and
discovery of common repeated musical patterns (ii) the evaluation of the musical
creativity exhibited through the assessment of the musical improvisations in terms of
a newly constructed creativity model and (iii) the application of contrast data
mining, i.e. the identification of differences of repeated musical patterns found in a
corpus, with respect to another one.
In order to realise the above goal, a computational model has been introduced,
designed and implemented. This work and its results are presented in this thesis.
Main subject category:
Technology - Computer science
Keywords:
Computational Music Analysis, Children Improvisations, Music Data Mining
Number of references:
243
