A Similarity Network for Correlating Musical Structure to Military Strategy

TL;DR

The paper addresses the challenge of understanding music structure by proposing an AI-driven, MFCC-based Music Clips Correlation Network (MCCN) to systematically analyze music perception and its parallels with military strategy. It trains MCCNs from music clips and compares their network properties to four military system networks (RTN, RAN, SOS, BA-NW-C2NM) using metrics such as Average Path Length, Network Diameter, Graph Density, Modularity, and Clustering Coefficient, along with a weighted dissimilarity $D_{weighted}^{mus, mil}$. Empirical results from ~2000 clips across ~60 war-film soundtracks reveal that offensive music tends to exhibit hierarchical structure and aligns most with the SOS network, while defensive music shows broader similarity to BA, SOS, and RAN. The work offers a cross-domain, systemic perspective on music perception and education, demonstrating how network analysis can illuminate structural similarities between musical composition and military planning, with potential applications in aesthetics education and multimodal perception research.

Abstract

Music perception, a multi-sensory process based on the synesthesia effect, is an essential component of music aesthetic education. Understanding music structure helps both perception and aesthetic education. Music structure incorporates a range of information, the coordination of which forms the melody, just as different military actions cooperate to produce a military strategy. However, there are a few ways for assessing music perception from the perspectives of system operation and information management. In this paper, we explore the similarities between music structure and military strategy while creating the Music Clips Correlation Network (MCCN) based on Mel-frequency Cepstral Coefficients (MFCCs). The inspiration comes from the comparison between a concert conductor's musical score and a military war commander's sand table exercise. Specifically, we create MCCNs for various kinds of war movie soundtracks, then relate military tactics (Sun Tzu's Art of War, etc.) and political institutions to military operations networks. Our primary findings suggest a few similarities, implying that music perception and aesthetic education can be approached from a military strategy and management perspective through this interdisciplinary research. Similarly, we can discover similarities between the art of military scheming and the art of musical structure based on network analysis in order to facilitate the understanding of the relationship between technology and art.

Figures (6)

• Figure 1: The techniques and steps used to study the "systematicness " of music are presented in this paper.
• Figure 2: MCCN Visualizing process for song "550W/Moss". (a) The initial graph of MCCN, which visualizes the music data. (b) The updated graph of MCCN by using the Yifan Hu layout algorithm. (c) The militarized music network, which ranks the nodes and Lines by betweenness centrality. We first update the music network layout by using an algorithm. And then, we calculate betweenness centrality for all nodes and rank the nodes, which is the process from (b) to (c). Nodes that are darker and bigger than others have more relations and similarities with other clips in this music. Also, we can regard the core nodes as the commander of this music because this clip has the most betweenness centrality in the whole music. We can observe that the network for this piece of music exhibits a distinct hierarchical structure. This hierarchy is visually evident through variations in node depth, size, and arrangement. Furthermore, within this musical composition, three segments can be considered relatively significant, as they correspond to the three largest nodes in the network. Additionally, these three segments in the composition exhibit a higher degree of connectivity with other segments. For example, they share similarities in terms of pitch or melody with other segments.
• Figure 3: (a) Random Tree Network and (b) Random Apollo Network which both include 50 nodes.
• Figure 4: Graph (a) is the System of System Network and BA-NW-C2NM Network is the (b) graph, which both include 50 nodes.
• Figure 5: (a) and (b) are offensive MCCNs, while (c) and (d) are defensive MCCNs. Offensive MCCNs have obvious hierarchy, while nodes in defensive MCCNs almost have the same betweenness centrality.