數位孿生系統(Digital Twin)藉由模擬真實環境的動態運作，在運作過程中不斷累積與生成大量具有時間與空間屬性的資料，匯聚成多元、跨域及多維度結構的時空資料。即使數位孿生機制被期許提供強大的分析能力，如何透過知識觀點由大量資料中彰顯地理現象之時空特色，以協助決策之制定仍是極大的挑戰。地理資訊系統(Geographic Information System, GIS)於描述與整合時空資料中具有顯著優勢，可在數位孿生系統中扮演關鍵角色，以地理資訊系統為核心技術，強調時空資料之視覺化呈現成為有效簡化資訊、凸顯資料特性與潛藏趨勢的重要技術之一。然而現行之視覺化技術應用仍高度依賴操作者的領域知識與軟體操作經驗，缺乏智慧的視覺化技術選擇與自動化的操作能力。對於不具專業背景的使用者而言，從時空資料中獲取與探索具價值的分析結果仍具高度門檻。為了提升視覺化展示的準確性與降低設計之門檻，過去研究提出多種圖表分類準則與選擇策略，但這些準則大多缺乏時間與空間性的相關描述，在面對龐大且多變的時空資料時，仍難以因應數位孿生各類成果之複雜度。
前述討論問題之關鍵因素在於成功的視覺化設計機制涉及需求分析、資料內容、判斷處理、視覺化策略選擇及展示內容設計等各類知識，要同時具有這些知識及軟體操作能力需要大量與長期的培養與訓練。本研究以建立視覺化技術選擇規則為目標，透過資料特性分析、視覺化需求分類，最後透過與大型語言模型整合，並將選擇規則結合地理資訊系統軟體，實作為智慧之視覺化系統。本研究首先基於文獻回顧分析時空資料特性、依據時間性、空間性與主題性，歸納資料可能存在之特質。接著將常見之視覺化需求依據時間模式與目的分類，分別提出四類時間模式（單一時間、兩時間差、時序變化、趨勢預測）與四類視覺化目的（分佈、比較、組成、關係），作為視覺化情境設計之依據，並透過交叉分析，形成視覺化策略建議之基礎。另一方面，本研究亦彙整並分析常見之視覺化技術，並針對每項視覺化技術分析其適用資料條件、應用場景與限制。分析之結果結合前述資料特性與需求分類邏輯，歸納整理出一套多層級的視覺化技術選擇規則，以時間模式為首層分類依據，依序搭配視覺化目的與資料特性條件，推薦不同場景下適用之視覺化技術，且作為後續自動化執行程序之根基。為實現上述視覺化技術推薦策略，本研究進一步發展一套智慧製圖系統，實際以大型語言模型Mistral Large 2.1結合地理資訊系統軟體ArcGIS Pro進行開發。使用者可透過自然語言表達需求，系統便能解析語意、辨識視覺化需求類型，並依據資料特性推薦圖表，產生相應之視覺化成果。最終於系統測試中包含多種視覺化場景與資料類型，並測試系統因應語意模糊、同資料不同需求等情境下之處理能力，結果顯示系統具備一定的語意容錯性與推薦一致性，並可成功協助使用者完成視覺化任務。
本研究提出之視覺化技術選擇規則與實作系統兼具理論架構與實務可行性。由理論發展觀點，透過系統性建構資料特性、視覺化需求與視覺化技術三者之對應關係，可補足現有研究對於視覺化策略推薦準則與應用場景分類之不足；實務上則透過語言模型降低使用門檻，提供一套能理解自然語言、推薦及產製視覺化成果的智慧決策輔助工具，適合非GIS專業背景使用者之應用情境。基於此融入視覺化知識之系統架構，此系統可望為未來時空資料智慧視覺化發展提供可行的參考方向與實證基礎，並進一步支援數位孿生系統中龐大且複雜的時空資料視覺化需求，有助於輔助呈現模擬成果與支援決策功能。

Digital Twin systems simulate the dynamic operations of real-world environments, continuously generating vast amounts of data with temporal and spatial attributes. These datasets accumulate into complex, multi-dimensional big spatiotemporal data. The massive scale and heterogeneity of such data pose significant challenges for analysis, interpretation, and visualization. Geographic Information Systems (GIS) offer distinct advantages in managing and integrating spatiotemporal data, making them increasingly essential tools for digital twin applications. Leveraging GIS for data visualization has become a crucial means to simplify complex information, highlight data characteristics, and reveal hidden patterns. However, current visualization practices rely heavily on users’ domain expertise and lack standardized criteria for selecting appropriate visualization techniques or automated operational procedures. For non-expert users, extracting meaningful insights from spatiotemporal data remains a high-barrier task. To enhance both the accuracy and accessibility of visualization, prior studies have proposed various chart classification schemes and selection strategies. However, these often fail to address the temporal and spatial dimensions of data explicitly, making them inadequate for the complexities inherent in big spatiotemporal data and digital twin scenario.

This study aims to establish a rule-based framework for selecting visualization techniques. It begins with analyzing data characteristics and categorizing visualization purposes, then integrates this logic with a language model to develop an intelligent visualization system within a GIS environment. First, based on literature review, the study classifies spatiotemporal data into three primary dimensions: temporal, spatial, and thematic, thereby identifying key data traits. It further categorizes common visualization needs into four temporal patterns and four analytical purposes, providing a foundation for defining visualization scenarios. On the other hand, this study compiles and analyzes widely used visualization techniques, detailing their applicable data requirements, use contexts, and limitations. By integrating knowledge of visualization techniques with visualization needs, it proposes a multi-level decision rule for visualization technique selection. This rule begins with temporal patterns, followed by visualization purposes, and finally considers spatial and thematic data attributes, allowing for the recommendation of the appropriate visualization method under different scenarios. To implement this strategy, this study develops an intelligent visualization system that combines a large language model (Mistral Large 2.1) with GIS software (ArcGIS Pro). Users can submit their visualization requests using natural language, which the system interprets to identify visualization needs, evaluate the characteristics of the provided data, recommend suitable visualization technique, and generate corresponding outputs. The system was tested across various data types and visualization scenarios, including cases involving ambiguous language and different demands on the same dataset. The results demonstrate the system’s semantic tolerance and recommendation consistency, enabling users to complete visualization tasks successfully.

The proposed framework and system contribute both theoretically and practically to the field. Theoretically, this study constructs a structured mapping between data characteristics, user needs, and visualization techniques, addressing the current lack of comprehensive chart recommendation rules and scenario classification. Practically, by integrating natural language processing, the system lowers entry barriers and provides a decision-support tool capable of understanding user intentions, recommending appropriate charts, and producing visualization outputs, which particularly beneficial for users without GIS expertise. This system holds promise as a scalable and practical reference model for future intelligent spatiotemporal visualization systems. Moreover, it directly supports the visualization demands of digital twin systems, which rely on the accurate and efficient presentation of complex spatiotemporal simulations to enhance decision-making and model interpretation.

林信助. (2008). 開放式時空地理資料描述基本架構之研究 [碩士論文, 國立成功大學]. 臺灣博碩士論文知識加值系統. https://hdl.handle.net/11296/dp8nhy
社會經濟資料服務平台. (2025). https://segis.moi.gov.tw/STATCloud/QueryInterface?
政府資料開放平臺. (2018). 臺中市犯罪統計資訊. https://data.gov.tw/dataset/84151
政府資料開放平臺. (2020). 滿15歲以上現住人口之教育程度－依鄉鎮市(88-101年). https://data.gov.tw/dataset/128926
政府資料開放平臺. (2024). 113年現住人口數按性別及年齡分. https://data.gov.tw/dataset/169795
環境部. (2025). 空氣品質指標(AQI)(歷史資料). https://data.moenv.gov.tw/dataset/detail/AQX_P_488
Abela, A. (2006). Choosing a good chart. https://extremepresentation.typepad.com/blog/2006/09/choosing_a_good.html
Andrienko, G., Andrienko, N., Demsar, U., Dransch, D., Dykes, J., Fabrikant, S. I., Jern, M., Kraak, M.-J., Schumann, H., & Tominski, C. (2010). Space, time and visual analytics. International Journal of Geographical Information Science, 24(10), 1577-1600. https://doi.org/10.1080/13658816.2010.508043
Atluri, G., Karpatne, A., & Kumar, V. (2018). Spatio-Temporal Data Mining: A Survey of Problems and Methods. ACM Comput. Surv., 51(4), Article 83. https://doi.org/10.1145/3161602
Beale, L. (2017). Insights fot ArcGIS. https://www.esri.com/content/dam/esrisites/en-us/media/pdf/data-visualization-options-in-insights-english-3.0.pdf
Camoes, J. (2013). A classification of chart types. http://excelcharts.com/classification-chart-types/
Card, S. K., & Mackinlay, J. (1997, 21-21 Oct. 1997). The structure of the information visualization design space. Proceedings of VIZ '97: Visualization Conference, Information Visualization Symposium and Parallel Rendering Symposium,
Chi, E. H. (2000, 9-10 Oct. 2000). A taxonomy of visualization techniques using the data state reference model. IEEE Symposium on Information Visualization 2000. INFOVIS 2000. Proceedings,
Demchenko, Y., Laat, C. d., & Membrey, P. (2014, 19-23 May 2014). Defining architecture components of the Big Data Ecosystem. 2014 International Conference on Collaboration Technologies and Systems (CTS),
Dominguez, D., Pantoja, O., Pico, P., Mateos, M., Alonso-Almeida, M. d. M., & González, M. (2020). Panama Papers' offshoring network behavior. Heliyon, 6(6), e04293. https://doi.org/https://doi.org/10.1016/j.heliyon.2020.e04293
Dübel, S., Rohlig, M., Schumann, H., & Trapp, M. (2014). 2D and 3D presentation of spatial data: A systematic review. 2014 IEEE VIS International Workshop on 3DVis (3DVis), 11-18. https://doi.org/10.1109/3DVis.2014.7160094
Ed Huai-Hsin, C., & Riedl, J. T. (1998, 19-20 Oct. 1998). An operator interaction framework for visualization systems. Proceedings IEEE Symposium on Information Visualization (Cat. No.98TB100258),
Esri. (2025a). Create Space Time Cube By Aggregating Points (Space Time Pattern Mining). https://pro.arcgis.com/en/pro-app/latest/tool-reference/space-time-pattern-mining/create-space-time-cube.htm
Esri. (2025b). Emerging Hot Spot Analysis (Space Time Pattern Mining). https://pro.arcgis.com/en/pro-app/latest/tool-reference/space-time-pattern-mining/emerginghotspots.htm
Esri. (2025c). How Emerging Hot Spot Analysis works. https://pro.arcgis.com/en/pro-app/latest/tool-reference/space-time-pattern-mining/learnmoreemerging.htm
Esri. (2025d). How Generate Desire Lines works. https://desktop.arcgis.com/es/arcmap/latest/extensions/business-analyst/create-spider-diagrams-desire-lines.htm
Fang, T. B., & Lu, Y. (2011). Constructing a Near Real-time Space-time Cube to Depict Urban Ambient Air Pollution Scenario. Transactions in GIS, 15(5), 635-649. https://doi.org/https://doi.org/10.1111/j.1467-9671.2011.01283.x
Fegeas, R. G., Cascio, J. L., & Lazar, R. A. (1992). An Overview of FIPS 173, The Spatial Data Transfer Standard. Cartography and Geographic Information Systems, 19(5), 278-293. https://doi.org/10.1559/152304092783762209
Fowler, R. A., Fletcher, T., Fischer, W. A., 2nd, Lamontagne, F., Jacob, S., Brett-Major, D., Lawler, J. V., Jacquerioz, F. A., Houlihan, C., O'Dempsey, T., Ferri, M., Adachi, T., Lamah, M. C., Bah, E. I., Mayet, T., Schieffelin, J., McLellan, S. L., Senga, M., Kato, Y., . . . Bausch, D. (2014). Caring for critically ill patients with ebola virus disease. Perspectives from West Africa. Am J Respir Crit Care Med, 190(7), 733-737. https://doi.org/10.1164/rccm.201408-1514CP
Hägerstraand, T. (1970). WHAT ABOUT PEOPLE IN REGIONAL SCIENCE? Papers in Regional Science, 24(1), 7-21. https://doi.org/https://doi.org/10.1111/j.1435-5597.1970.tb01464.x
Hedley, N., Drew, C., Arfin, E., & Lee, A. (1999). Hagerstrand Revisited: Interactive Space-Time Visualizations of Complex Spatial Data. Informatica (Slovenia), 23, 155–168.
Holtz, Y., & Healy, C. (2018). from Data to Viz. https://www.data-to-viz.com/
Johnson, B., & Shneiderman, B. (1991, 22-25 Oct. 1991). Tree-maps: a space-filling approach to the visualization of hierarchical information structures. Proceeding Visualization '91,
Kang, Y., Cho, N., & Son, S. (2018). Spatiotemporal characteristics of elderly population’s traffic accidents in Seoul using space-time cube and space-time kernel density estimation. PLOS ONE, 13(5), e0196845. https://doi.org/10.1371/journal.pone.0196845
Keim, D. A. (2002). Information visualization and visual data mining. IEEE Transactions on Visualization and Computer Graphics, 8(1), 1-8. https://doi.org/10.1109/2945.981847
Keller, P. R., & Keller, M. M. (1993). Visual cues: practical data visualization (Vol. 2). IEEE Computer Society Press Los Alamitos, CA.
Kraak, M.-J. (2003a). Geovisualization illustrated. ISPRS Journal of Photogrammetry and Remote Sensing, 57(5), 390-399. https://doi.org/https://doi.org/10.1016/S0924-2716(02)00167-3
Kraak, M.-J. (2003b). The space-time cube revisited from a geovisualization perspective. 21st international Cartographic Conference, ICC 2003,
Lake, M. R., Burggraf, D., Trninic, M., & Rae, L. (2004). Geography mark-up language: foundation for the geo-web (Vol. 1). John Wiley & Sons.
Li, Z., & Ning, H. (2023). Autonomous GIS: the next-generation AI-powered GIS. International Journal of Digital Earth, 16(2), 4668-4686. https://doi.org/10.1080/17538947.2023.2278895
MacEachren, A. M., & Kraak, M.-J. (2001). Research Challenges in Geovisualization. Cartography and Geographic Information Science, 28(1), 3-12. https://doi.org/10.1559/152304001782173970
Mackinlay, J. (1986). Automating the design of graphical presentations of relational information. ACM Trans. Graph., 5(2), 110–141. https://doi.org/10.1145/22949.22950
Mansourian, A., & Oucheikh, R. (2024). ChatGeoAI: Enabling Geospatial Analysis for Public through Natural Language, with Large Language Models. ISPRS International Journal of Geo-Information, 13(10).
Mo, C., Tan, D., Mai, T., Bei, C., Qin, J., Pang, W., & Zhang, Z. (2020). An analysis of spatiotemporal pattern for COIVD-19 in China based on space-time cube. J Med Virol, 92(9), 1587-1595. https://doi.org/10.1002/jmv.25834
Nyerges, T. L. (1991). Geographic Information Abstractions: Conceptual Clarity for Geographic Modeling. Environment and Planning A: Economy and Space, 23(10), 1483-1499. https://doi.org/10.1068/a231483
Riehmann, P., Hanfler, M., & Froehlich, B. (2005, 23-25 Oct. 2005). Interactive Sankey diagrams. IEEE Symposium on Information Visualization, 2005. INFOVIS 2005.,
Schneider, M., & Weinrich, B. E. (2004). An abstract model of three-dimensional spatial data types Proceedings of the 12th annual ACM international workshop on Geographic information systems, Washington DC, USA. https://doi.org/10.1145/1032222.1032234
Shneiderman, B. (1996, 3-6 Sept. 1996). The eyes have it: a task by data type taxonomy for information visualizations. Proceedings 1996 IEEE Symposium on Visual Languages,
Standardization, I. O. f. (2002). Geographic information — Temporal schema (ISO Standard No. 19108:2002). In.
Standardization, I. O. f. (2019a). Date and time — Representations for information interchange (ISO Standard No. 8601-1:2019). In.
Standardization, I. O. f. (2019b). Geographic information — Spatial schema (ISO Standard No. 19107:2019). In.
Swabish, J., & Ribecca, S. (2014). The Graphic Continuum. https://policyviz.com/2014/09/09/graphic-continuum/#google_vignette
Tang, A. Y., Adams, T. M., & Lynn usery, E. (1996). A spatial data model design for feature-based geographical information systems. International journal of geographical information systems, 10(5), 643-659. https://doi.org/10.1080/02693799608902102
Team, F. T. V. J. (2015). Visual Vocabulary. https://github.com/Financial-Times/chart-doctor/tree/main/visual-vocabulary
Technology, N. I. o. S. a. (1994). Spatial Data Transfer Standard (SDTS). In.
Tory, M., & Moller, T. (2004, 10-12 Oct. 2004). Rethinking Visualization: A High-Level Taxonomy. IEEE Symposium on Information Visualization,
Tufte, E. R., & Graves-Morris, P. R. (1983). The visual display of quantitative information (Vol. 2). Graphics press Cheshire, CT.
Usery, E. L. (1996). A Feature-Based Geographic Information System Model. Photogrammetric Engineering & Remote Sensing, 62, 833-838.
Wickham, H. (2014). Tidy Data. Journal of Statistical Software, 59(10), 1 - 23. https://doi.org/10.18637/jss.v059.i10
Xu, D., Zhang, Q., Ding, Y., & Zhang, D. (2021). Spatiotemporal Pattern Mining of Drought in the Last 40 Years in China Based on the SPEI and Space–Time Cube. Journal of Applied Meteorology and Climatology, 60(9), 1219-1230. https://doi.org/https://doi.org/10.1175/JAMC-D-21-0049.1
Yin, Z., Zhang, C., Goldberg, D. W., & Prasad, S. (2019). An NLP-based Question Answering Framework for Spatio-Temporal Analysis and Visualization Proceedings of the 2019 2nd International Conference on Geoinformatics and Data Analysis, Prague, Czech Republic. https://doi.org/10.1145/3318236.3318240
Yuan, H., Chen, B. Y., Li, Q., Shaw, S.-L., & Lam, W. H. K. (2018). Toward space-time buffering for spatiotemporal proximity analysis of movement data. International Journal of Geographical Information Science, 32(6), 1211-1246. https://doi.org/10.1080/13658816.2018.1432862
Zhang, Y., Wei, C., Wu, S., He, Z., & Yu, W. (2023). Geogpt: Understanding and processing geospatial tasks through an autonomous gpt. arXiv preprint arXiv:2307.07930.
Zlatanova, S. (2006). 3D Geometries in Spatial DBMS. In A. Abdul-Rahman, S. Zlatanova, & V. Coors (Eds.), Innovations in 3D Geo Information Systems (pp. 1-14). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-540-36998-1_1