Publications

Computational Humanities Research 2024

Abstract |

Color symbolism is considered a critical element in art and literature, yet determining the relationship between colors and words has remained largely subjective. This research presents a systematic methodology for quantifying the correlation between language and color. We utilize text-based image search, optical character recognition (OCR), and advanced image processing techniques to establish a connection between words and their corresponding color distributions in the CIELch color space. We generate a color dataset based on human cognition, and apply it for analysis of the literary works of poets associated with Imagism and Black Arts Movements. This helps uncover the characteristic color patterns and symbolic meanings of the movements with enhanced objectivity and reproducibility in literature research. Our work has the potential to provide a powerful instrument for a systematic, quantitative examination of literary symbolism, filling in the gaps in prior analyses and facilitating novel investigations of thematic aspects using color.

NetSci 2024

Abstract |

Legislations, much like academic publications, contain citations and references to statutes, cases, regulations, or secondary sources, enabling readers to locate and verify the cited authority. Citations can be broadly divided into two types: the Consistent type where the two legislations (the citing one and the cited one) are in accordance with each other, and the Contrary type, where they are in conflict or opposite in meaning for various reasons. An excess of contrary citations is an indication of a high level of inconsistencies in the corpora of legislations, an undesirable feature for an extensive set of rules that are enforced on all members in a country or jurisdiction who are at risk of being penalized or punished for violating it. From the network science view, these two types of connections can be modeled as positive and negative edges in a signed network. For this study we collected the statutes of the Republic of Korea in its entirety, then performed natural language processing (NLP) to determine the signs of the citations. We then measured the structural balance of the network to quantify its coherence and robustness. This study provides insights into Korea’s recent legislative trends–notably the rise in special laws for expediency–and introduces a novel, systematic quantitative approach to the field of Law and Development, an interdisciplinary field that seeks to diagnose and formulate ideal legal frameworks for societal advancement, by identifying pivotal statutes and their collective roles in the legal landscape.

한국서양음악이론학회 2024

Abstract |

제2회 한국음악저작권협회 논문공모전 수상집 ‘생성형 AI와 음악 저작권’ 부문 장려

Abstract |

Proceedings of the 25th International Society for Music Information Retrieval Conference (ISMIR), 2024

Abstract | This study aims to measure the similarity of melodies objectively using natural language processing (NLP) techniques. We utilize Mel2word which is a melody tokenization method based on byte-pair encoding to facilitate the semantic analysis of melodies. In addition, we apply two word weighting methods: the modified Tversky measure for word salience and the TF-IDF method for word importance and uniqueness, to better understand the characteristics of each melodic element. We validate our approach by comparing song vectors calculated from an average of Mel2Word vectors to the ground truth in 103 cases of music copyright infringement, sourced from an extensive review of legal documents from law archives. The results demonstrate that the proposed approach is more in accordance with court rulings and perceptual similarity.

AES (Audio Engineering Society) 155th convention 2023

Abstract | This paper presents a comprehensive analysis of the evolution of research topics in audio engineering over the past four decades. This study’s aim is to examine the change of research trends by analyzing abstracts from the Journal of the Audio Engineering Society (JAES). Using the g-DMR topic modeling technique, the authors identified 16 key research trends from 2,038 abstracts from the JAES. These findings provide an analysis of how sound engineering has evolved in relation to the progress of media technology, showing the rise and fall of the research keywords such as spatial hearing, loudspeaker response and measurement, spatial reproduction, music information retrieval (MIR), etc. The authors hope that this historical analysis will help reflect on the society’s development and look to the future, with plans to expand the analysis with more comprehensive data covering all the AES publications including its international convention proceedings.

CNA (International Conference on Complex Networks & Their Applications) 20th 2023

Abstract | Prose poetry, written in a prose format free from the formal constraints of traditional poetry, is distinct from regular prose. Its proliferation is understood as a feature of modernity. Despite the prominence of prose poetry in Korean literature, its inherent nature—straddling between traditional poetic elements and prose—has posed challenges for scholars. This study, adopting the “distant reading” approach, seeks to quantify and characterize the nuances of Korean prose poetry by comparing its semantic structures with regular prose. We contrasted major 21st-century Korean prose poems with news articles from Naver—selected for their fact-based nature. Both were tokenized into morphemes using the kiwi library, and represented as networks where tokens served as nodes. Connections between these tokens were defined by a window size of 2 and quantified using fasttext word embeddings. Results highlighted distinct differences: prose poetry showed a greater variability in word connection strengths and denser word clusters, suggesting its inherent ambiguity. In contrast, regular prose exhibited stronger connections and more direct paths between words, indicating a more straightforward structure.

CMMR (International Symposium on Computer Music Multidisciplinary Research) 2023

Abstract | In this study we present a network analysis of the communities of artists based on sampling. We construct a bipartite network between the artists who perform the sampling and the samples, then detect communities of the artists and the samples. We find that sample-based music has a clear community structure where each community features artists (nodes) with high centralities, allowing us to determine its musical style. We also define and visualize the similarities between communities representing distinct generations to observe how sample-based musical styles have evolved or been “handed off” to the posterity. This study not only enhances our understanding of sampling-based music, but also presents a novel application of network community structure to a creative enterprise such as music.

CMMR (International Symposium on Computer Music Multidisciplinary Research) 2023

Abstract | Musical composition can be viewed as an act of conditional problem solving, the realization of musical ideas by arranging notes spatially and tempo- rally. The resulting creations may constitute the unique style of the composer. In this paper we focus on how chord voicing – the expression of chords by choosing and stacking musical notes – has evolved in western classial piano music using large-scale music data sets. Our results shows that the level and variety of voicing novelty have increased throughout history. We also find that some composers exhibit a high level of voicing novelty due to the utilization of innovative pitch class sets, while others actually have pushed the boundaries of voicing with tradi- tional pitch class sets. This study helps us to probe the emergence of expression of musical style on note level and to understand the evolutionary pattern of note arrangements.

NetSci (International School and Conference on Network Science) 2023

Abstract | The study of history is a fascinating field of inquiry that seeks to uncover and understand the patterns of human actions under varying political, economic, or natural conditions, as well as the consequences that flow from those actions. While the complexity of the continually evolving social system makes it unlikely that a set of clean and concise “laws of history” will ever be found, the increasing availability of data on past human actions and social conditions has opened the avenue for “quantitative history”, which is the history-side analog of computational social science. In recent times, patterns of accomplishments by individuals in science, culture, and athletics have garnered much attention. This is likely driven by the fact that these feats are often widely admired by connoisseurs and aficionados for their excellence or perceived value to society, and they produce concrete objects in the form of publications, artworks, figures (numbers), etc. that can be systematically collected, Figure. The interaction network of three royal figures (King Tanjong and his uncles Suyang and Anp’yŏng) and bureaucrats during Suyang’s Revolt of 1493. Suyang successfully overthrew Tanjong, while Anp’yŏng failed in his own revolt at the same time and was eventually executed. The edges in the network represent co-appearance in the same article in the Annals of Chosŏn during a two-month period surrounding the revolt, whereas the node colors represent the post-revolt status of the bureaucrats: Blue means purged and red means decorated by Suyang. This shows how political events and governance are correlated with the social network of major figures: A simple visual inspection suggests that the purged tend to be close to Anp’yŏng, and the decorated to Suyang. The inset in the lower left shows it in more quantitative terms: nodes with large eigenvector centrality–indication of close association with the central nodes, possibly partisanship and alliance–tend to be either decorated (Suyang’s allies) or purged (Anp’yŏng’s allies), whereas those with high betweenness centrality–indication of being intermediaries between central nodes rather than being partisan–do not show such a tendency. organized, and analyzed. However, an area of human activity that has an even more significant impact on society is politics. Politics can be broadly defined as ‘the activities associated with the governance of a country’ or ‘an authoritative allocation of values’. The importance of politics and governance in the functioning and evolution of a highly complex system such as a society cannot be overstated. A large-scale study of individuals’ actions and accomplishments in the field that spans the entire duration of a historical regime is all the more valuable because it offers a glimpse into how the society functioned and, when the governance is in a crisis, what kind of dynamics evolve that threaten its existence. In this presentation, we introduce and analyze the official government records from the Chosŏn dynasty (1392-1910 CE) that ruled the Korean peninsula and its inhabitants. These records are widely recognized as an exemplary specimen of meticulous, comprehensive, and reliable historic data. Utilizing modern quantitative methods on detailed information on Chosŏn’s bureaucracy and its people, we aim to find clues to the following questions to help us obtain a better understanding of how social systems evolve and decay: What type of talents did Chosŏn’s vaunted bureaucracy recruit into itself? How did the bureaucrats’ careers unfold over their lifetimes, and how did they correlate with major historical events? And what kinds of signals can we extract about Chosŏn’s fate from these data? The practical benefits importance of governance cannot be overstated — functioning government, the opposite of anarchism, is considered necessary for an orderly society and active dissolution of discord and conflict between various social constituents. Characterizing the patterns of governance and bureaucracy during the Chosŏn dynasty could help us understand how the society functioned, and when governance is in crisis, what kind of dynamics evolve that threaten its existence. Ultimately, this research can help us obtain a better understanding of how social systems evolve and decay, and thus help us prepare for a better future.

NetSci (International School and Conference on Network Science) 2023

Abstract | In the big data era, distant reading techniques in literary analysis allow us to see broader patterns traditionally hidden in close reading. Our study applies this approach to color imagery research in literature, which has been limited by the subjectivity of color assignment in previous studies. Using a corpus of works by English Imagist poets, we automated the assignment of Basic Color Terms to words, followed by a statistical analysis and topic modeling to connect theme, work, and color.

Our findings provide insights into each poet’s unique use of color words and potential thematic-color relationships. Our multi-layered network analysis, incorporating assortativity, community detection, and robustness metrics, reveals patterns, associations, and critical nodes in these relationships. This approach paves the way for nuanced and objective color imagery research in large literary corpora, potentially offering broader insights across different literary movements with further data collection and analysis.

NetSci (International School and Conference on Network Science) 2023

Abstract | Word embedding is a Natural Language Processing (NLP) technique for finding the vector representation of a word in terms of other words that reflects the syntactic or semantic context of word usage, enabling us to quantify the association between two words. In this study we use the technique to examine how the authors’ “ideas” – the perception of the subject materials – have changed over time in Science Fiction. We employ Dynamic Bernoulli Embeddings (DBE), a type of word embedding that hypothesizes that embedding vectors shift through the vector space over time. DBE produces different embedding vectors for the same word in different time periods, enabling us to explore the transitions in word meanings and associations. We can visualize this using the egocentric network of the word of interest. In the Figure, we show two egocentric networks of the word “world” in the 1910’s and the 1970’s where the words are connected when the cosine similarity between the embedding vectors is 0.5 or larger. The associated words that forming a densely connected community (e.g. a clique) around the ego show how the context or the word meaning may differ in different era. We anticipate this approach to aid us in the quantitative evaluation of literary works.

NetSci (International School and Conference on Network Science) 2023

Abstract | Despite an elevated level of interest in understanding human creativity in recent years, there have not been many network-inspired studies on how creative thoughts–novel and useful combinations of concepts–manifest themselves in the cognitive conceptual space; the quantification of lexical creativity [1] or the examination of the semantic networks of creative people [2] have yet to explain the temporal dynamics or the rise of creativity. As a conduit for human knowledge transfer and diffusion, the structure and characteristics of the underlying cognitive network of concepts may significantly affect human creativity, warranting an exploration and analysis of it [3] by investigating how humans connect different concepts. Here we introduce the KAIST Knowledge Graph (KKG), a free-text knowledge graph dataset constructed through voluntary participation by the university’s members (students, staff, and faculty). Composed of 817 nodes and 1 495 edges with detailed descriptions in natural language, KKG differs from classical existing knowl- edge graphs such as Freebase and Wikidata in that it is a direct reflection of the cognitive and metaphoric relationships inside people’s thoughts, not dictionary-based ones. Starting from the initial node set prescribed by the KAIST president, the participants were asked to add new nodes into the network inspired by the existing ones and make connections among them. The properties of each node or edge include name, description in free-text form, time created, user id who created it, and related images if available. We find that the network exhibits hubs and a strong community (see Figure), implying non-trivial but structured associations between concepts, however, the associations show little correlation with their associations in the meaning-based Con- ceptNet[4]. This implies that understanding the mechanism behind those associations made inside the brain that transcend meaning would be key understanding human creativity.

Journal of the Korean Physical Society

Abstract | Musical sampling is a composition technique of popular music where one borrows elements from existing recordings to produce new songs. The sampled music is thus deeply related to the newly created music based on it. We can, therefore, surmise that the sampling practice of an artist reflects the characteristics of the subgenre or the music community which the artist belongs operates in and belongs to, implying that the sampling relationships can help us understand the origin and evolution of many different styles of sample-based music. In this study, we present a complex network analysis of the communities of artists connected via sampling relationships. We establish an artist–sample bipartite network of artists who perform the sampling, and the songs that are the subjects of sampling. The detection of communities composed of artists and songs demonstrates that the sample-based musical scene has a clear community structure where each community features artists with high centrality that allows us to identify the musical styles of the community. While this study focuses on the understanding of sample-based music that forms the basis of an overwhelming majority of contemporary popular musical paradigms, we believe this framework is general enough to be applied to many other creative fields that involve referencing of existing works.

NetSci (International School and Conference on Network Science) 2022

Abstract | How can we explore the relationship between literary works? With recent advances in Natural Language Processing (NLP), it is becoming increasingly feasible to identify connections between written works based on content itself. This `machine reading’ enables us to characterize the structure of the network in literature in large scale. In this study we analyze the topical subjects in SF (science fiction) appearing the classical pulp magazine Amazing Stories. To this end we employ `topic modeling’ to identify the underlying topics of the SF novels, then form the network of SF stories based on their topical similarity. An analysis of the network reveals the type of topics in SF and how they are used in conjunction. We also define measures of an author’s distinctiveness and diversity of topic usage, which allow us to understand the writer’s literary characteristics. Based on this work we anticipate a more robust quantitative framework for the evolution of ideas and the transmission of messages in stories that can further illuminate many network aspects of literary works.

NetSci (International School and Conference on Network Science) 2022

Abstract | Creativity is the foundation of all artistic and intellectual innovations, but it is not yet well charac- terized. As creativity is increasingly believed to be the ability to make novel connections between concepts in many enterprises, exploring the actual patterns of connections in creative artifacts is gaining attention. While some works exist on semantic memory networks and linguistic cre- ativity, there have been fewer works on the network patterns of musical and artistic creativity, a topic of intrigue especially in the current era of computer-generated music and art driven by advances in AI and deep learning. Here we present the patterns of melodic creativity in the classical piano works composed between the Baroque and the modern periods. After computing the network-based creativity scores of the works reflecting novelty and influence [1], we measure various topological properties of the network of compositional elements and study how they cor- relate to one another. We find that clustered melodic transitions implies high creativity, whereas longer distances and higher modularity–the existence of distinct groups connected preferentially internally–imply low creativity, showing that creativity is manifest in the ready ability to scale the whole space of compositional elements and establish new connections, supporting the Flatter Association Hierarchies of Mednick’ hypothesis [2] and paralleling previous comparative studies on linguistic creativity [3,4,5]. Our paper demonstrates the possibility of an intriguing and novel understanding of artistic creativity from the perspective of network science.

JOURNAL OF THE KOREAN PHYSICAL SOCIETY, v.80, no.6, pp.533 – 542, 2022-03

Abstract | Cultural creations such as movies can be said to be the products of the times, influenced by successful predecessors and in turn influencing those that come afterwards. In this work we use network science and computational social science-inspired methods to analyze a large-scale data set of movies focusing on how they have changed and evolved over time using a comprehensive movie-meme association data to construct the network of movies that form the timeline of the history of cinema via the evolution of genres, and the rise and fall of prominent sub-genres. We also identify the impactful movies that were harbingers to popular memes that dominate a given period of time and from which certain genres form and grow. Finally, we measure how the impact of movies correlates with the experts’ and the public’s assessment to understand the conditions for success and further development. We believe this work showcases how network science can be used to understand the evolution of important components of society.

PHYSICAL REVIEW RESEARCH, v.3, no.2, 2021-06

Abstract | Given the rapidly evolving landscape of linguistic prevalence, whereby a majority of the world’s existing languages are dying out in favor of the adoption of a comparatively fewer set of languages, the factors behind this phenomenon have been the subject of vigorous research. The majority of approaches investigate the temporal evolution of two competing languages in the form of differential equations describing their behavior at a large scale. In contrast, relatively few consider the spatial dimension of the problem. Furthermore while much attention has focused on the phenomena of language shift-the adoption of majority languages in lieu of minority ones-relatively less light has been shed on linguistic coexistence, where two or more languages persist in a geographically contiguous region. Here, we study the geographical component of language spread on a discrete medium to monitor the dispersal of language species at a microscopic level. Language dynamics is modeled through a reaction-diffusion system that occurs on a heterogeneous network of contacts based on population flows between urban centers. We show that our framework accurately reproduces empirical linguistic trends driven by a combination of the Turing instability, a mechanism for spontaneous pattern-formation applicable to many natural systems, the heterogeneous nature of the contact network, and the asymmetries in how people perceive the status of a language. We demonstrate the robustness of our formulation on two datasets corresponding to linguistic coexistence in the northwestern part of Spain and the southern part of Austria.

PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS, v.565, 2021-03

Abstract | Sports audiences’ sense of excitement originates from a multitude of factors such as the uncertainty in the outcome of a game and the expectation of winning streaks. The uncertainty factor can be said to be maximized when the game is tied. At the same time, a tie represents the antipode to the ultimate goal-to win-of the contestants and the wishes of their loyal fans. A tie therefore encourages the contestants to continually adapt to the situations and strategize to break it, leading to an even more dynamic and engrossing gameplay. A key to understanding this phenomenon starts from the characteristic dynamics of ties and scoring events in sports games. Here we analyze the complete data from a full season of the National Basketball Association (NBA), the professional basketball league of the United States and Canada, to find the patterns of scoring and ties and how they correlate with the interactive nature of sports, and show how they differ from traditional simple random models based on cruder summary statistics that can show their insufficiencies on fine details of gameplay. Given the social and economic significance of such enterprises, these types of findings will prompt the much-needed developments in detailed modeling of sports based on actual data.

EPJ DATA SCIENCE, v.9, no.1, 2020-01

Abstract | Recent advances in the quantitative, computational methodology for the modeling and analysis of heterogeneous large-scale data are leading to new opportunities for understanding human behaviors and faculties, including creativity that drives creative enterprises such as science. While innovation is crucial for novel and influential achievements, quantifying these qualities in creative works remains a challenge. Here we present an information-theoretic framework for computing the novelty and influence of creative works based on their generation probabilities reflecting the degree of uniqueness of their elements in comparison with other works. Applying the formalism to a high-quality, large-scale data set of classical piano compositions-works of significant scientific and intellectual value-spanning several centuries of musical history, represented as symbolic progressions of chords, we find that the enterprise’s developmental history can be characterised as a dynamic process composed of the emergence of dominant, paradigmatic creative styles that define distinct historical periods. These findings can offer a new understanding of the evolution of creative enterprises based on principled measures of novelty and influence.

PLOS ONE, v.14, no.12, 2019-12

Abstract | Human communication is invariably executed in the form of a narrative, an account of connected events comprising characters, actions, and settings. A coherent and well-structured narrative is therefore essential for effective communication, confusion caused by a haphazard attempt at storytelling being a common experience. This also suggests that a scientific understanding of how a narrative is formed and delivered is key to understanding human communication and dialog. Here we show that the definition of a narrative lends itself naturally to network-based modeling and analysis, and they can be further enriched by incorporating various text analysis methods from computational linguistics. We model the temporally unfolding nature of narrative as a dynamical growing network of nodes and edges representing characters and interactions, which allows us to characterize the story progression using the network growth pattern. We also introduce the concept of an interaction map between characters based on associated sentiments and topics identified from the text that characterize their relationships explicitly. We demonstrate the methods via application to Victor Hugo’s Les Miserables. Going beyond simple, aggregate occurrence-based methods for narrative representation and analysis, our proposed methods show promise in uncovering its essential nature of a highly complex, dynamic system that reflects the rich structure of human interaction and communication.

PHYSICAL REVIEW E, v.99, no.6, 2019-06

Abstract | Mechanisms of pattern formation-of which the Turing instability is an archetype-constitute an important class of dynamical processes occurring in biological, ecological, and chemical systems. Recently, it has been shown that the Turing instability can induce pattern formation in discrete media such as complex networks, opening up the intriguing possibility of exploring it as a generative mechanism in a plethora of socioeconomic contexts. Yet much remains to be understood in terms of the precise connection between network topology and its role in inducing the patterns. Here we present a general mathematical description of a two-species reaction-diffusion process occurring on different flavors of network topology. The dynamical equations are of the predator-prey class that, while traditionally used to model species population, has also been used to model competition between antagonistic features in social contexts. We demonstrate that the Turing instability can be induced in any network topology by tuning the diffusion of the competing species or by altering network connectivity. The extent to which the emergent patterns reflect topological properties is determined by a complex interplay between the diffusion coefficients and the localization properties of the eigenvectors of the graph Laplacian. We find that networks with large degree fluctuations tend to have stable patterns over the space of initial perturbations, whereas patterns in more homogenous networks are purely stochastic.

PLOS ONE, v.13, no.9, 2018-09

Abstract | Painting is an art form that has long functioned as a major channel for the creative expression and communication of humans, its evolution taking place under an interplay with the science, technology, and social environments of the times. Therefore, understanding the process based on comprehensive data could shed light on how humans acted and manifested creatively under changing conditions. Yet, there exist few systematic frameworks that characterize the process for painting, which would require robust statistical methods for defining painting characteristics and identifying human’s creative developments, and data of high quality and sufficient quantity. Here we propose that the color contrast of a painting image signifying the heterogeneity in inter-pixel chromatic distance can be a useful representation of its style, integrating both the color and geometry. From the color contrasts of paintings from a large-scale, comprehensive archive of 179 853 high-quality images spanning several centuries we characterize the temporal evolutionary patterns of paintings, and present a deep study of an extraordinary expansion in creative diversity and individuality that came to define the modern era.

IEEE ACCESS, v.6, pp.29,098 – 29,108, 2018-06

Abstract | Human behaviors are known to spread through social contact. The diffusion process on social networks has also been leveraged to understand the spread of undesirable contagion. The contagion of malicious or even criminal behaviors in online social networks is just beginning to attract attention. Here, we study the social contagion problem of cheating behavior found in the massively multiplayer online roleplaying game (MMORPG) that provides a lifelike environment with rich and realistic user interactions. Because cheating users boast an abnormal thus conspicuous degree of success, it has a strong chance of being noticed by their friends and leading them to cheat themselves. To detect and prevent cheating, it is beneficial to understand this dynamic as a contagion problem. In this paper, we show the existence of the contagion of cheating. We then explore various possible social reinforcement mechanisms after introducing several factors to quantify the effect of social reinforcement on the contagion and analyze the dynamics of bot diffusion in an extensive user interaction log from a major MMORPG.

ADVANCES IN COMPLEX SYSTEMS, v.21, no.3-4, 2018-05

Abstract | In the modern era where highly-commodified cultural products compete heavily for mass consumption, finding the principles behind the complex process of how successful, “hit” products emerge remains a vital scientific goal that requires an interdisciplinary approach. Here, we present a framework for tracing the cycle of prosperity-and-decline of a product to find insights into influential and potent factors that determine its success. As a rapid, high-throughput indicator of the preference of the public, popularity charts have emerged as a useful information source for finding the market performance patterns of products over time, which we call the on-chart life trajectories that show how the products enter the chart, fare inside it, and eventually exit from it. We propose quantitative parameters to characterize a life trajectory, and analyze a large-scale data set of nearly 7,000 songs from Gaon Chart, a major weekly Korean Pop (K-Pop) chart that covers a span of six years. We find that a significant role is played by nonmusical extrinsic factors such as the established fan base of the artist and the might of production companies in the on-chart success of songs, strongly indicative of the commodified nature of modern cultural products. We also review a possible mathematical model of this phenomenon, and discuss several nontrivial yet intriguing trajectories that we call the “Late Bloomers” and the “Re-entrants” that appear to be strongly driven by serendipitous exposure on mass media and the changes of seasons.

PLOS ONE, v.13, no.2, 2018-02

Abstract | The quality of life for people in urban regions can be improved by predicting urban human mobility and adjusting urban planning accordingly. In this study, we compared several possible variables to verify whether a gravity model (a human mobility prediction model borrowed from Newtonian mechanics) worked as well in inner-city regions as it did in intra-city regions. We reviewed the resident population, the number of employees, and the number of SNS posts as variables for generating mass values for an urban traffic gravity model. We also compared the straight-line distance, travel distance, and the impact of time as possible distance values. We defined the functions of urban regions on the basis of public records and SNS data to reflect the diverse social factors in urban regions. In this process, we conducted a dimension reduction method for the public record data and used a machine learning-based clustering algorithm for the SNS data. In doing so, we found that functional distance could be defined as the Euclidean distance between social function vectors in urban regions. Finally, we examined whether the functional distance was a variable that had a significant impact on urban human mobility.