Comparative Pattern Analysis of Cretan Folk Songs
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
∗
Comparative Pattern Analysis of Cretan Folk Songs
Darrell Conklin
Department of Computer Science and Artificial Intelligence
Universidad del Paı́s Vasco
San Sebastián, Spain
IKERBASQUE: Basque Foundation for Science
Bilbao, Spain
conklin@ikerbasque.org
Christina Anagnostopoulou
Department of Music Studies
University of Athens
Athens, Greece
chrisa@music.uoa.gr
Abstract
This paper reports on data mining of Cretan folk songs for distinctive patterns. A
pattern is distinctive if it occurs with higher probability in a corpus as compared to
an anticorpus. A small set of Cretan folk songs was collected, organized using a small
knowledge base of classes, and mined using distinctive pattern discovery methods.
Several highly distinctive and confident patterns emerge.
keywords: ethnomusicology, data mining, folk song analysis, pattern discovery, classifi-
cation
1 Introduction
In recent years there is a renewed interest in folk song analysis partly driven by increasing
interests in cultural heritage, and also by advances in music informatics methods. The
∗
This is the author’s version of the work. It is posted here by permission of ACM for your personal
use. Not for redistribution. Please cite the definitive version which was published at MML’ 10, 3rd
International Workshop on Machine Learning and Music, October 25, 2010, Florence, Italy,
pages 33–36.
1ability to make predictions, from music content, of song properties such as region, dance
type, tune family, instrumentation, modality, and social function is an important part of
the management of large corpora. Music data mining methods play a key role in building
predictive models for folk song classification.
The folk music of Crete, like all traditional Greek music, can be divided primarily into
dance and non-dance (which are further subdivided into “Tavla” or “songs of the table”
and “road” songs). There are various types of songs belonging to each of these categories.
For dance songs, the most common types found in Crete are Kontylies and Pentozalis,
Malevisiotis, Syrtos and Sousta. The songs that are not danced are Rizitika, Tambahaniotika,
and other table songs, and then various function type songs, such as lament songs, wedding
songs, and lullabies.
Although most of these types of songs are found throughout the island, each area has its
own characteristic musical style. The music of the West of Crete is quite distinct from the
music of the East, although there are also many common songs known to the whole island.
Therefore, in stylistic analysis of this music, it is necessary to look not only for patterns that
distinguish between song types, but also how the music of each region differs.
This paper reports on some initial results obtained with a small collection of folk tunes
from various villages across Crete, exploring the hypothesis that there may exist patterns of
correlation between origin, function and type of song on the one hand, and melodic patterns
on the other.
As an analysis method we have chosen the MGDP method [4, 5] which discovers patterns
that are distinctive and maximally general in a corpus. This paper will discuss the encoding
of a collection of pieces into a knowledge base for analysis, describe the application of the
MGDP method to the corpus, report on some interesting patterns found, and finally present
some ideas for future work.
22 Method
This section describes the data preparation and data mining methods employed. The data
mining method finds patterns that are over-represented in a positive set (called the corpus)
with respect to a background set of pieces (called the anticorpus).
2.1 Corpus
Three main sources were used for the Cretan folk songs: the first one was the collection of
transcriptions done by Samuel Baud-Bovy in 1953-54 [3], published by the Laographic Music
Archives of Melpo Merlie in Athens. It contains old and unknown songs which were recorded
and transcribed in various villages around the island. The second source was by Peristeris
[8], as found in the Amargiannakis archives at the Academy of Athens and the University of
Athens. The third one was a collection of well-known songs, transcribed for lyra playing [2].
Finally, a few pieces were added from the transcriptions done by Theodossopoulou [9].
Songs from printed scores were encoded using the Sibelius and Finale software, and
exported to symbolic midi format for computational analysis. Following encoding of the
pieces, a knowledge base was developed to describe the classes of all pieces. Four broad
categories were chosen: song type, song hypertype, area, and hyperarea (see Figure 1).
Encoded songs were classified into the song types classification specified by Baud-Bovy
[3]. Hyperarea classification was according to a map of Crete, and hypertype according to
the classification of Greek folk music specified by Amargiannakis [1] and Baud-Bovy [3] and
inspection of song lyrics.
While preparing the corpus and classifying the songs, several interesting points came up
for discussion, and decisions had to be taken:
• as mentioned above, there are song types found throughout the island, such as the new
year song (kalanda). These songs were therefore placed in all classes of area (and
3hyperarea);
• there are also songs which might have been recorded at a specific village (e.g., a small
village of Chania), but these songs are sung all through the Western part of the island:
These songs were thus added to both rethymno and chania areas;
• for a few songs, although the song type was clear, there was no indication in the sources
regarding its area. Therefore, in our classification, these pieces are not considered part
of any class for either area or hyperarea mining;
• finally, some songs were noted as simply being Tavla songs, without any further song
type classification, and from the music it was not possible to deduce a more specific
description. In our opinion this meant that they should be treated as not known for
the song type category, though they still participated in the song hypertype class
tavla.
A total of 106 songs was in our final collection. Figure 1 depicts the categories along with
their classes and the number of pieces in each class.
2.2 MGDP discovery
A pattern is a sequence of event features. A rich set of features, including rhythmic ones,
can be ascribed to events [4], but for this study only melodic intervals are used due to the
small corpus size. A piece x instantiates a pattern P , written P (x), if the pattern occurs
(possibly multiple times) in the piece: if the components of the pattern are instantiated by
successive events in the piece.
A pattern P subsumes (is more general than) a pattern Q if all instances of Q are also
instances of P : if ∀xQ(x) ⇒ P (x) is valid (true in all possible corpora). For example, the pat-
tern [+3] subsumes the pattern [+3, +1], which in turn subsumes, for example, [+2, +3, +1]
4hyperarea hypertype
west (64) east (47) tavla (61) dance (51)
song type
sousta (5) wedding (14) city song (7) pentozalis (11) kontylies (17) lament (8) nanourisma (12) epic song (1) malevisiotis (1) rizitiko (8) syrtos (17) kalanda (1)
area
chania (30) rethymno (29) herakleion (12) lassithi (35) syria (4)
Figure 1: The categories and classes used in this study, along with the number of pieces in
each class.
5and [+3, +1, −4].
To rank discovered patterns it is possible to partition a collection of pieces into an explicit
anticorpus (denoted ), contrasting it with the analysis corpus (denoted ⊕). A distinctive
pattern is one that is frequent and sufficiently over-represented in the corpus as compared
to the anticorpus. An intuitive way to measure this is according to the relative empirical
probability of a pattern in the corpus and anticorpus:
def p(P |⊕)
∆(P ) = (1)
p(P | )
where
def
p(P |⊕) = c⊕ (P )/n⊕
def
p(P | ) = c (P )/n
where c⊕ (P ) (c (P )) is the total count of a pattern P in the corpus (anticorpus), and n⊕
(n ) is the number of pieces in the corpus (anticorpus).
A pattern P is called a maximally general distinctive pattern (MGDP) if it is distinc-
tive (with at least a specified minimum ∆(P )), and if there does not exist a more general
(subsuming) pattern that is also distinctive. Thus the set of all MGDP is the top border of
the virtual pattern subsumption taxonomy such that all patterns above the border are not
distinctive.
Patterns are ranked primarily by distinctiveness (Equation 1). Additionally, for every
distinctive pattern the confidence of the association rule P ⇒ ⊕ is computed. This is the
probability of the class given the pattern:
c⊕ (P )
def
p(⊕|P ) = , (2)
c(P )
6where c(P ) is the piece count of pattern P in the entire piece collection, which may differ
from the quantity c⊕ (P ) + c (P ) due to overlap between classes.
3 Results
For each category hyperarea, area, song hypertype, and song type the MGDP set is
found by in turn considering each class as the corpus ⊕ and all remainding classes as the
anticorpus . For example, for the area category, considering the class chania, the classes
rethymno to syria (see Figure 1) are taken together as the anticorpus.
For all experiments, melodic intervals patterns with ∆(P ) ≥ 3 were sought. The mini-
mum p(P |⊕) used was 20% of the corpus. Also, very sparse classes syria, kalanda, male-
visiotis, and epic song were not considered as positive corpora as it is possible for even
the low 20% threshold to be satisfied by just a single piece.
Table 1 shows a subset of patterns found, presenting 3 patterns for each category that are
highly distinctive with ∆(P ) ≥ 5, with piece count c⊕ (P ) ≥ 5. Several patterns also have
high confidence (Equation 2). In the following we discuss one pattern from each category.
In the dance song hypertype the pattern [+4, −4] is found in all types of dances, through-
out the island. The instances of this pattern are found in fast isochronous passages. Four
instances, taken from different types of dances are shown in Figure 2 (top row). In most
instances the second note occurs on a relatively strong beat.
In the west hyperarea, an interesting, frequent, and high confidence pattern [−4, +2, −3]
was found. The pattern occurs in nearly 25% of the west pieces, and does not occur at all in
east pieces. This pattern is also distinctive, but to a lesser degree, in the rethymno area.
In diatonic terms, this represents a drop of a M3, an ascending M2, followed by a descending
m3, for instance, E-C-D-B or A-F-G-E. Figure 2 (middle row) shows seven instances of the
pattern: the second and third are found in the same song, but with a different surrounding
7class ⊕ pattern P c⊕ (P )/n⊕ p(⊕|P ) ∆(P )
area
lassithi [−7, +4] 7/35 0.88 14.2
lassithi [−2, −1, +1, −1, +1, −1] 7/35 0.88 14.2
rethymno [−4, +2, −3] 10/29 0.71 6.6
song type
syrtos [−5, −2] 5/17 1.00 ∞
kontylies [+3, −7] 5/17 0.83 25.0
kontylies [+1, +2, +3] 5/17 0.83 25.0
hyperarea
west [−4, +2, −3] 14/64 1.00 ∞
east [+1, +2, −2, +2, −2] 13/47 0.81 5.9
east [−2, +2, −2, −1, +1] 11/47 0.79 5.0
song hypertype
dance [+4, −4] 16/51 0.94 19.1
dance [−4, +2, +2, −4] 11/51 0.92 13.2
dance [+4, +1, +2] 19/51 0.90 11.4
Table 1: Patterns found in the Crete folk song corpus, organized by category.
Figure 2: Instances of the dance pattern [+4, −4], west pattern [−4, +2, −3], and syrtos
pattern [−5, −2].
8context. Thus the phrase is varied while this core pattern remains intact.
For the syrtos song type, an interesting and high confidence pattern [−5, −2] was found.
The pattern occurs in no other song types. In diatonic terms, this pattern represents a
descending P4 followed by a descending M2, for instance, A-E-D or E-B-A. Figure 2 (bottom
row) shows five instances of the pattern. The rhythmic and metrical aspects of the instances
varies considerably, indicating the abstraction provided by the melodic interval pattern.
4 Conclusions
Though the amount of data is small (just over 100 pieces), some interesting findings have
emerged. First, in the hyperarea category, all highly distinctive patterns are within the
dance class (data not shown). Also, two maximally distinctive patterns (not occurring in
any other classes) were found, in the syrtos song type, and the west hyperarea.
The method and knowledge base developed is extensible and readily admits adding fur-
ther Crete folk songs. In addition, a novelty here was the allowance for pieces to participate
in multiple classes.
In this experiment the MGDP algorithm was applied by taking each class in turn as
the corpus; alternatively it is entirely possible to group classes together (e.g., wedding and
lament) to form a compound corpus. Concurrently, the data mining algorithm could easily
be modified to use a background song hierarchy so, for example, the pattern [−4, +2, −3]
would not have been reported twice (in both area and hyperarea) but only the most
distinctive occurrence.
It is clear that more songs are needed, especially to more thoroughly populate the classes
of the area and type categories. Even with a threshold of 20%, for small classes it is too
easy to reach this threshold, and this is borne out by the preponderance of type patterns
from small classes such as rizitiko, city song, sousta, and lament (data not shown: in
9Table 1, the total output was first filtered to retain only the frequent distinctive patterns).
There are some general considerations to bear in mind with the analysis of Cretan folk
music (and folk music in general): first, as mentioned above, transcription of the songs
deviates from the live performance [6]. Second, these songs are found in various villages and
places, with minor variations; the transcriptions available to us did not notate the variations.
Third, there are songs that go beyond the boundaries of their origin, and they can be found in
other areas with possible variations. Fourth, a song of the table might become at a different
instance a song of the road, and even a dance song [7] if different lyrics are sung. Finally,
in Cretan music especially, a lot of these songs are performed by various artists, and each
lyra artist has a unique style of playing, which in this case means adding small characteristic
turns and motives. Information about performer might be taken into account but would
require more data collection for distinctive pattern analysis.
Acknowledgments
This work was partially funded by the John Latsis Foundation (www.latsisfoundation.org),
through the project grant entitled “Comparative computational analysis of Cretan music”,
in the call “Meletes 2009”. Many thanks also to N. Poulakis, R. Pylarinos, and A. Pikrakis
for encoding and discussions on the corpus.
References
[1] G. Amargiannakis. Gia mia morphologia tou ellinikou dimotikou tragoudiou (towards
a morphology of the greek traditional song). Technical report, Department of Music
Studies, University of Athens, 1994.
10[2] E. Andreoulakis and S. Petrakis. Music and Mantinades in Crete. Philippos Nakas Music
Publications, Athens, Greece.
[3] S. Baud-Bovy. Mousiki Katagrafi stin Kriti 1953-1954 (Music recordings in Crete 1953-
1954). Folk Music Archives Melpo Merlie, Athens, Greece, 2006.
[4] D. Conklin. Discovery of distinctive patterns in music. Intelligent Data Analysis,
14(5):547–545, 2010.
[5] D. Conklin. Distinctive patterns in the first movement of Brahms’s String Quartet in C
Minor. Journal of Mathematics and Music, 4(2):85–92, 2010.
[6] M. Dragoumis. I Paradosiaki mas Mousiki (Our Traditional Music) Vol. 2. Folk Music
Archives Melpo Merlie, Athens, Greece, 2009.
[7] M. Rombou-Levidi. Psifides horou ston evro. In L. Droulia and L. Liavas, editors,
Mousikes tis Thrakis: Mia diepistimoniki Proseggisi. Syllogos oi Filoi tis Mousikis, Evros,
1999.
[8] G. Spuridakis and S. Peristeris. Greek Folk Songs. Academy of Athens Publications,
Athens, Greece, 1968.
[9] E. Theodossopoulou. Oi skopoi stin Elliniki Paradosiaki Moysiki (Tunes in Greek Tradi-
tional Music). PhD thesis, Department of Music Studies, University of Athens, Athens,
Greece, 2003.
11You can also read
