|
For a comprehensive and intrinsic study on this topic, I am going to reference this article:
https://towardsdatascience.com/the-data-science-of-k-pop-understanding-bts-through-data-and-
a-i-part-1-50783b198ac2
And I am just going to cornerstone BTS' singular and distinguishable factors because that is the
major topic. And I do not intend to digress a lot.
───────────────────────
The objective of this study and answer is not intended to proselytize the idea of BTS-pop nor is
it intended to denigrate any other groups mentioned.
This was written with the sole aim of providing an answer to the question of “why do some
people use the term ‘BTS-pop’?”
I reiterate: the opinion is “The idea of ‘BTS-pop’ is feasible”. A possibility. Not the complete truth.
Contention 1: Speechiness
Prerequisites
:
For an arrangement of 11 acoustic qualities in BTS’ music, Spotify’s API has been inspected.
This gives a mathematical measurement of the numeric level of the tracks’ acoustic-ness,
danceability, instrumental-ness, etc.
The calculations are solely based on Spotify’s internal algorithm for dissimilation of various
mathematical properties within BTS’ music.
Kpop artists: Based off ‘Top 20 Artists’ list.
Number of Kpop tracks for research: 2,673 soundtracks.
Observation
:
The above chart delineates that BTS’ music had the highest level of *speechiness.
*Speechiness: The detection of spoken words in a track. Vocals, to be precise.
The value is almost tripling the value of speechiness from the other artists.
For example, BigBang and BTS level of speechiness:
0:158097 [BTS] > 0.108219 [BB]
Hence, 0.158097 [BTS] - 0.108219 [BB] = 0.049806
Accurately, BTS’ music possesses 1.4608987331244x times of BigBang’s level of speechiness
if we divide the numericals.
Contention 2: A.I. Modelling
Prerequisite
:
As it is productively easier for machines to memorize algorithms for musical attributions to
differentiate one music artist from the other. The analyser has built a simple machine classifier
that learns those 11 musical features, predicting if they are indeed from BTS or not.
Ensemble Models Used: LightGBM, Gradient Boosting and Random
Forest.
Model’s AUC: 0.9
Target Label 1: Corresponding to BTS’ music
Target Label 0: Not corresponding
Mode: Combat Class Imbalance for Oversampling.
Observations
:
Kpop Vs. BTS:
[This is a SHAPE value chart. Used for demonstrating mathematical calculations of specific
features that aid to predict if the track belongs to BTS or not. Based on those 11 features.]
How to read the data:
● More impact on the model prediction is seen by the level of high features on the y-axis.
● The magnitude of positive and negative impact on model output is shown by how spread
out the dots are from the center.
● Positive influence means influence toward BTS prediction, and negative influence
toward the other side.
● Colors blue and red in the chart indicate the value of a feature.
● The red dots represent the higher levels of speechiness.
● And the blue represents the lower values of this feature.
The most important feature in the model prediction is speechiness. The higher the level of
speechiness, the more that song influenced the machines in predicting it to be by BTS among
all Kpop artists.
Contention 3: BTS vs. American Pop Artists
Prerequisite:
The same models of study are used while taking American artists such as Kelly Clarkson, One
Direction, and Bruno Mars into consideration to differentiate BTS’ music.
Observation:
BTS’ value of speechiness is again twice that of other Pop artists.
For example, BTS and Christina Aguilera
0.158097 [BTS] > 0.108485 [CA]
And so, 0.158097 [BTS] - 0.108485 [CA] = 0.049612
Accurately, 1.4573166797253x times that of Christina Aguilera's value of spechiness in her
music .
The SHAP value chart:
Again, the most important feature in the model prediction is the level of speechiness conducted
by the blue dots. Hence, the prediction again favors BTS’ music.
Contention 4: Tempo
BigBang & iKon’s tempo
: 125 ~ 90 respectively
BTS’ tempo
: 180 ~ 35
BTS’s music is far more distributed in the beats. Indicating that BTS’ music is much more
diverse. The rhythm and speed being fast and slow paced are equally diverse.
Which in turn explains that BTS’ music is indeed the most diversified of all.
Conclusion
While each and every member of the ‘BTS-pop’ community is not a music analytic or data
analyser of sorts, these properties within BTS’ music might seem idiosyncratic to them.
The audience may have caught up on the uniqueness of BTS’ music which stirred in them an
urge to classify BTS as a complete other genre.
While at the same time, disintegrating them from Kpop and more so even the mainstream.
Eventually giving rise to the notion of ‘BTS-pop
Do'stlaringiz bilan baham: |
