aiqbench analysis
This study applies unsupervised graph learning techniques to analyze semantic structures within K-pop lyrics, demonstrating how different embedding representations influence the formation of cultural communities. While the dataset is culturally specific, the core methodology relies on multilingual MiniLM models and network partitioning algorithms to map textual similarity without human labeling. The work highlights technical trade-offs in representation choice and graph construction when organizing large-scale unstructured text corpora.
Semantic Communities and Boundary-Spanning Lyrics in K-pop: A Graph-Based Unsupervised Analysis
arXiv:2602.12881v2 Announce Type: replace-cross Abstract: We examine how lyric representation shapes computational similarity using four 384-dimensional multilingual MiniLM representations of 7,891 K-pop songs by 1,456 artists from charts spanning 2000 to October 2023. Whole-song truncated embeddings provide the strongest standardized separation between neighbouring and random song pairs (1.995), compared with 1.931 for chunked documents, although chunked representations show greater adjusted Rand agreement and line-based representations yield higher modularity. Using whole-song embeddings, we construct a weighted cosine nearest-neighbour graph ($k=30$). A representative Louvain partition identifies 21 operational communities, without a corresponding robust plateau in the Markov Stability analysis. AI-assisted examination of selected lyrics and terms suggests overlapping patterns of memory, separation, longing, self-reflection, attraction, and performance. Community membership is associated with character-script composition (Cram\'er's $V=0.294$; permutation $p=0.001$). High-participation songs exhibit broader cross-community connectivity, while participation has weak negative associations with embedding-relative novelty. These findings show how representation choices and textual characteristics influence the network organization of a cultural corpus, subject to limitations from truncation, graph construction, and the absence of independent human semantic validation.
Abstract:We examine how lyric representation shapes computational similarity using four 384-dimensional multilingual MiniLM representations of 7,891 K-pop songs by 1,456 artists from charts spanning 2000 to October 2023. Whole-song truncated embeddings provide the strongest standardized separation between neighbouring and random song pairs (1.995), compared with 1.931 for chunked documents, although chunked representations show greater adjusted Rand agreement and line-based representations yield higher modularity. Using whole-song embeddings, we construct a weighted cosine nearest-neighbour graph ($k=30$). A representative Louvain partition identifies 21 operational communities, without a corresponding robust plateau in the Markov Stability analysis. AI-assisted examination of selected lyrics and terms suggests overlapping patterns of memory, separation, longing, self-reflection, attraction, and performance. Community membership is associated with character-script composition (Cramér's $V=0.294$; permutation $p=0.001$). High-participation songs exhibit broader cross-community connectivity, while participation has weak negative associations with embedding-relative novelty. These findings show how representation choices and textual characteristics influence the network organization of a cultural corpus, subject to limitations from truncation, graph construction, and the absence of independent human semantic validation.
Social and Information Networks (cs.SI); Computation and Language (cs.CL)
arXiv:2602.12881 [cs.SI]
(or arXiv:2602.12881v2 [cs.SI] for this version)
https://doi.org/10.48550/arXiv.2602.12881 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 266 words
Related coverage
U-CECE: A Universal Multi-Resolution Framework for Conceptual Counterfactual Explanations
arXiv · Oct 8, 2026
OSMGraphCLIP: Learning Global Location Representations from OpenStreetMap Graphs
arXiv · Jun 9, 2026
Manage agents, tools and skills at scale with AWS Agent Registry
AWS · Aug 31, 2026
Graph-centric agentic intelligence
Amazon Science · Oct 1, 2026