User Preference Maps: Quantifying the Built Environment

Abstract

The built environment in which we live holds the potential to provide life experiences that allow pedestrians to observe, feel, learn, and grow through their surroundings in everyday urban spaces. If a city offers opportunities for careful observation and exploration according to users’ preferences, it will become more appealing to many people. This study selected Midtown, New York, as the research site and collected a total of seven datasets based on 30 intersections in the area. The data, categorized into three main areas—activity, comfort, and natural elements—were evaluated, visualized, and restructured using a path exploration algorithm to produce a final user-based map. For this, 3D modeling software Rhino version 7, visual programming tool Grasshopper, and Grasshopper verion 2023 plugin programs were used. The final result included 3D route information, quantitative measurement data, and multidimensional visual materials. This approach presents an alternative to traditional route navigation based on uniform criteria and, through data-driven design, is believed to ultimately enhance walkability, activate urban spaces, and contribute to the development of sustainable cities. The scope of related research can further expand as the targets, duration, and methods of data collection continue to evolve and as case studies in various cities increase.

1. Introduction

1.1. Research Background

Navigation systems such as Google Maps 6.13 version and Kakao Maps 5.20 version, which we use regularly through smartphones, guide users along the same routes without offering much diversity in route selection based on user preferences. While the shortest route might be the only option in an emergency requiring quick escape, for pedestrians walking through a city in their daily lives, various surrounding factors could influence their choice of route beyond just the shortest distance [1,2,3,4,5,6,7,8,9,10,11,12,13]. A city has the potential to offer its inhabitants experiences that allow them to observe, learn, feel, enjoy, and grow. If a city can provide opportunities for users to carefully observe and explore its features according to their preferences, it will become more attractive to a larger audience [14,15,16,17]. In other words, pedestrians can feel a sense of comfort while viewing iconic landmarks, enjoy the cityscape through open spaces, and learn about local culture by walking down streets filled with famous restaurants. However, current navigation systems do not function as tools for providing personalized experiences.

The built environment of a city is defined as the environment encompassing buildings, transportation systems, open spaces, and land uses that influence humans [18,19,20]. As such, it is intricately connected to various fields, including architecture, urban planning, and transportation. Many studies have actively addressed how people perceive their surroundings and how these surroundings influence human behavior, as well as their social and economic impacts [21,22]. Scholars like Jan Gehl and Kevin Lynch have closely examined methods for activating urban spaces by linking them to the built environment and considering the relationship between urban spaces and individual users [23,24]. Pedestrian experiences hold significant value as a tool for urban exploration due to the high connectivity of various surrounding factors and the freedom of movement they offer. However, numerous studies on pedestrian route selection have been conducted by traffic planners, focusing on distance and time [25,26,27,28,29]. This study aims to qualitatively visualize the built environment based on a quantitative evaluation of various urban data and connect this assessment with the optimal path algorithm to propose a pedestrian-centric urban exploration map.

Amidst the rapid urbanization taking place today, new theories related to architectural and urban planning are emerging. Among them, New Urbanism presents the idea that enhancing walkability can revitalize a city and ultimately help design a sustainable urban environment [30]. This user-centered navigation map aims to allow people living in cities to experience various aspects of urban life, thereby improving walkability and contributing to the creation of sustainable cities [31,32,33].

1.2. Research Purpose

This study focuses on the area of Midtown Manhattan, from W40th St to W35th St, facing Bryant Park and the New York Public Library. It proposes a route exploration map with the southwestern corner of the area as the start point and the northeastern corner as the end point. This location was selected because it comprises dense office buildings and public facilities like Bryant Park and the New York Public Library, offering both functional and spatial diversity. To utilize various elements of the built environment, such as open street landscapes, landmark buildings, natural elements, building uses, and surrounding density, urban data for this street area are collected and evaluated, ultimately generating a route map. The final proposed navigation map aims to overcome the limitations of existing uniform route guidance tools by allowing users to experience urban routes based on their preferences. The goal of this study is to propose an interactive, two-way communication navigation map that contributes to sustainable urban design.

1.3. Research Stages

This study follows the process outlined in

Table 1. Process and tools.

Process 1	Process 2	Process 3	Process 4
Data Mining	Categorization and Site Modeling	Simulation, Evaluation, and Visualization	Application of Algorithm and Map Visualization
Postman web version 2024, MS Excel, Crome	Rhino, Grasshopper	Rhino, Grasshopper	Rhino, PyCharm, Grasshopper

to quantify elements of the built environment and develop a new interactive navigation map based on this dataset.

The main stages of this study are as follows. First, urban data related to elements of the built environment are collected. The locations for data collection are listed in

Table 2. Raw data information.

Name of Data	Raw Data Reference
Restaurant Density	https://www.yelp.com/ (accessed on 5 October 2024)
Building Density (Built-FAR)	https://www.nyc.gov/site/planning/data-maps/open-data/dwn-pluto-mappluto.page (accessed on 5 October 2024)
Visibility (Openness)	https://grasshopperdocs.com/components/grasshopperintersect/isoVist.html (accessed on 5 October 2024)
Landmark View	https://www.nyc.gov/site/planning/data-maps/open-data/dwn-pluto-mappluto.page (accessed on 5 October 2024)
Tree Density	https://data.cityofnewyork.us/Environment/2015-Street-Tree-Census-Tree-Data/pi5s-9p35 (accessed on 5 October 2024)
Noise Complaints	https://data.cityofnewyork.us/Social-Services/311-Noise-Complaints/p5f6-bkga (accessed on 5 October 2024)
Sky Exposure	https://www.nyc.gov/site/planning/data-maps/open-data/dwn-pluto-mappluto.page (accessed on 5 October 2024)

, and tools such as Excel, Postman, PyCharm, and Google Chrome are used. Second, the collected data are categorized according to the characteristics of the built environment, as shown in

Table 3. Built environment characteristics, user preference categories, and related data.

Name of Data	User Preference Category	Built Environment
Restaurant Density	Activity	Building-Use
Building Density	Activity	Built-FAR 1, (Density of Population)
Noise Complaints Density	Activity, Comfort	Street Condition: Noise
Visibility (Openness)	Comfort	Street Condition: Open Space
Landmark View	Comfort	Street Condition: View
Tree Density	Nature	Street Condition: Nature (Tree)
Sky Exposure Ratio	Nature	Street Condition: Nature (Sky)

¹ (Building) Built—floor area ratio.

. The criteria for categorization are based on previous neighborhood environment studies, which have demonstrated significant impacts on users [34]. Following this, the site is selected, and the start and end points are specified on the site. Based on the route to the end point, 30 intersections are modeled. Third, the data are simulated and visualized using the 30 intersections as reference points. Rhino, a 3D modeling tool, and Grasshopper, a visual programming language, are used for this simulation. Fourth, the evaluated data are re-evaluated using an algorithm to generate the final map. In the next section, we will examine and discuss more detailed information regarding the tools used, research methodology, and specific data values in the process of data collection, data evaluation, map generation algorithms, and final visualization.

2. Materials and Methods

2.1. Theoretical Review

2.1.1. Data-Based Design Approaches

Recently, there have been various attempts in the fields of architecture and urban planning to utilize data and algorithms. Emerging design methodologies include genetic algorithm-based models that use Darwin’s theory of natural selection to optimize data results, geographic information systems (GISs) that utilize a variety of local data in conjunction with spatial data, evidence-based design (EBD), which applies patient data as a basis for designing hospitals, and data-driven design, which quantifies space using regional climate data and other evidence-based information [35,36,37,38]. In this study, a data-based design methodology is used to collect and evaluate various data from 30 intersections that connect the start point and the end point within the site. Based on these data, the study generates an optimal route navigation map tailored to each user.

2.1.2. Visual Programming Tool: Grasshopper

Grasshopper is a tool that uses visual programming and serves as an application for the 3D modeling software Rhino. It has been widely used for many years. As shown in Figure 1, outputs are generated by connecting components with predefined output values to input values. Grasshopper is intuitive and visually implements logic because it integrates with Rhino and offers various open-source applications online. Python version 3.10, as a visual programming language, is particularly accessible, allowing not only programming experts but also professionals from other fields, including architecture, to use it easily. In this study, Grasshopper was utilized for data collection, evaluation, and visualization, and a map generation algorithm was applied to derive the final route navigation map. While there is a drawback of the system becoming somewhat heavy when processing large amounts of data due to the integration with a 3D modeling environment, the ease of linking with various Python-based algorithms and the advantages of working within a 3D environment make it a tool that will continue to be essential for spatial professionals.

2.1.3. Grasshopper Add-On Program

Various applications for Grasshopper are available as open-source tools online. A representative add-on is Ladybug [39], which allows for solar energy analysis based on EnergyPlus climate data. In this study, multiple Grasshopper add-ons are used, including Ladybug for simulating sky visibility (sky exposure area factor), GHowl for converting latitude and longitude data into the 3D Rhino environment, and @it, which enables the use of shapefiles containing 3D building information in the modeling environment. These tools play a critical role in the data collection and simulation process.

2.2. Data Collection and Evaluation Methods

2.2.1. Site Location

The selected site is Midtown, Manhattan, a prominent urban street in New York. The start point is assumed to be the intersection of W35th St and 9th Ave in the southwest corner, while the end point is set at the intersection of W40th St and 5th Ave in the northeast corner. The site includes a mix of major office buildings and public facilities such as Bryant Park and the New York Public Library, making it a location that encompasses both spatial openness and functional diversity. The surrounding built environment offers a range of urban landscapes and ensures a comfortable pedestrian experience, which could broaden the scope of pedestrian interactions with different types of spaces. An overview of the site and a satellite image oriented with the top as north are shown in Figure 2.

The expected benefits of this study are as follows. First, the process of collecting, evaluating, visualizing, and applying seven types of data provides guidance on how to utilize spatial data in urban environments. Second, by creating maps that connect pedestrians with the built environment, this study serves as a case for how user experiences can be applied to products in urban, architectural, and transportation sectors. Third, offering new opportunities to observe urban spaces increases walkability, which in turn naturally enhances the sustainability of cities. This is fundamentally achievable through the bidirectional communication between the cities we live in and individuals.

2.2.2. Data Collection and Evaluation at the Site

In this study, data are collected from 30 points within the site, and it is broadly categorized into three aspects: density of activity, degree of comfort, and density of natural elements. These three aspects are quantified through data evaluation. These data are then formalized by a map generation algorithm, ultimately proposing a generative map that allows for route selection based on the preferences of pedestrians. The process is outlined as follows:

Data are collected from 30 points at the site, each within a 120-m radius, as shown in

Table 4. Segments in the site.

Points in the Target Area
Midtown Manhattan	30 Points in the Street
	Data collection radius = 120 m

(start point: s; end point: e).

• The collected data are divided into three categories.
• The categorized data are evaluated and visualized.
• The map generation algorithm is applied.
• The final map is generated and visualized.

Three Categories: 1. Density of Activity

The first element that significantly impacts the pedestrian environment is the density of activity [4,16,40,41,42]. In fact, the vibrancy felt in street spaces adds variety to an otherwise uniform urban landscape, creating a dynamic and lively city. This can be evaluated based on factors such as how many popular restaurants are in the area and how densely the surrounding buildings are concentrated. In this study, data from Yelp, a review app used by over 80 million people, are collected via the Yelp API V3 platform in JSON format, as shown in

Table 5. Data and file format.

Name of Data	Tool	Data File Format
Restaurant Density	Excel, Grasshopper	JSON or Text
Building Density	Grasshopper, @it, ghowl	SHP 1
Noise Complaints	Grasshopper, ghowl	CSV
Visibility (Openness)	Grasshopper, @it, ghowl	SHP 1
Landmark View	Rhino, Grasshopper, @it, ghowl	SHP 1, 3DM 2
Tree Density	Grasshopper, ghowl	CSV
Sky Exposure	Grasshopper, @it, ghowl, Ladybug	SHP 1

¹ Shape file/² Rhino file.

. Restaurant data (

Table A1. Postman and data mining.

Access Token	User Preference Category
Qxmn7lUjHGaXP1YfXc6eZKRVT4VyELFdxOGj1v6 VEHEIh7QRK6dzXeDZFisvJW	https://api.yelp.com/v3/businesses/search?latitude=40.752910&longitude=-73.989224&limit=50&radius=120 (accessed on 5 October 2024) (Postman Yelp V3 Environment)

)—including name, rating, latitude, and longitude—are extracted in Grasshopper (

Table 6. JSON file and evaluation in Grasshopper.

JSON File	Evaluation in Grasshopper

). Only restaurants with the highest ratings (3.5 stars or above) are filtered, and their numbers are measured. Building density in the surrounding area is derived by linking New York City’s Pluto data in shapefile format to Rhino using the @it component in Grasshopper. The built floor area ratio (FAR) data from Pluto are quantitatively measured, as shown in (

Table 7. SHP file and evaluation in Grasshopper.

SHP File	Evaluation in Grasshopper

). Visualization is used to help understand the data. If the built FAR is smaller than the legal FAR standard, it is visualized as “underbuilt”; if it is larger, it is visualized as “overbuilt”.

Three Categories: 2. Degree of Comfort

The second element among the three categories is the degree of comfort, which refers to how pleasant and comfortable pedestrians feel while experiencing the built environment. The degree of comfort also has a significant impact on the pedestrian environment [16,42,43]. Comfort can be interpreted as the feeling that a place is attractive or desirable [44], which is influenced by factors like surrounding noise and scenery. In this study, comfort is evaluated based on openness, the visibility of landmark buildings (key elements of desirable urban environments), and the density of noise complaints in the city. Open spaces in complex urban streets add variety to an otherwise monotonous cityscape. To measure the size of open spaces, a visibility simulation is conducted using New York City Pluto Data’s 3D shapefile and Grasshopper’s IsoVist component, as shown in

Table 8. IsoVist simulation and visualization in Grasshopper.

IsoVist Simulation	Visualization in Grasshopper

. The IsoVist component simulates visibility by establishing a start point based on the human eye level and creating sight lines to a target point that does not intersect with physical elements. The size of the visible area (blue background color) enclosed by these sight lines is measured, as illustrated in Figure 3.

In addition, the higher the visibility of a landmark building, the more attractive the street is likely to be from a pedestrian’s perspective. In this study, visibility is determined by measuring the sight line from the pedestrian’s start point to the highest point of the landmark building. If this sight line does not intersect with physical elements such as surrounding buildings, it is considered visible (

Table 9. Landmark View.

Start Point, Target Point, Sight Line

). For noise levels, data from New York City’s 311 Noise Complaints data are used to assess the degree of comfort. Types of noise, such as loud music, loud parties, and loud talking, as well as their locations (latitude and longitude), are extracted and analyzed. The number of noise-related events is then measured to make an assessment (

Table 10. Noise 311 Complaints type and ghowl visualization.

Noise Type	Visualization in Grasshopper

).

Three Categories: Density of Nature

The natural environment is one of the key elements that determine the quality of urban spaces and impact on the pedestrian environment [9,43,45]. The desire to be surrounded by nature and to enjoy it appears to be one of the most fundamental human rights. However, due to the large-scale buildings planned in urban areas, the ability to enjoy natural elements such as trees [46] and the sky can vary greatly depending on the location within the city. In this study, the density of nature is measured by determining the extent to which trees and the sky are visible from various points. The number of trees at each site is measured based on the latitude and longitude information from the NYC Tree Data, and the visible area of the sky from the pedestrian’s perspective is calculated by connecting the SyMask component from Grasshopper’s Ladybug add-on to the shape files of existing buildings (

Table 11. Shape file filtering (@it) and Ladybug SyMask analysis.

Shape File Import (@it Component)	Ladybug SyMask Analysis

).

2.3. Results of Data Evaluation and Visualization at the Site

As data application technologies advance and become diverse, abundant datasets become available; thus, methods for visualizing data effectively in urban architecture and transportation fields are being explored. In the past, data visualization was defined as the visual representation of quantitative data. However, this study seeks to explore multidimensional visualization by applying an information design model theory tailored to the characteristics of the urban architecture field [47,48,49]. This approach involves three key principles. First, it allows for computer-based quantitative interaction between the user and dimensional data. Second, it provides 3D spatial information alongside quantitative data. Third, it generates continuous visual materials that create an immersive environment for users. To represent the collected data and quantitative evaluation results according to these concepts, this study uses techniques that sequentially present quantitative dimensions, 3D spatial information, and continuous images to draw multidimensional interactions with users. The expression of multidimensional results serves not only as a tool for conveying quantitative data but also facilitates qualitative and aesthetic decision making necessary in the urban architecture field. Below,

Table 12. Restaurant density evaluation and visualization.

Number of Restaurants	Restaurants in the Circle	s, x1~x28, e: Restaurant Density (3.5~5 Stars from Yelp): Intersections in a Zigzag Order Starting from the Lower Left in the Location on the Map (Table 4)
Street Seg.	Num. Restaurant	Street Seg.	Num. Restaurant	Street Seg.	Num. Restaurant	Street Seg.	Num. Restaurant
s	8	x1	5	x2	3	x3	14
x4	10	x5	4	x6	5	x7	7
x8	8	X9	5	x10	13	x11	14
x12	8	x13	4	x14	4	x15	8
x16	13	x17	8	x18	17	x19	9
x20	12	x21	4	x22	7	x23	7
x24	1	x25	13	x26	12	x27	10
x28	6	e	6

,

Table 13. Building density evaluation and visualization.

Number of BuiltFAR	Points in the Circle	s, x1~x28, e: Building Density (Underbuilt/Overbuilt): Intersections in a Zigzag Order Starting from the Lower Left in the Location on the Map (Table 4)
Street Seg.	BuiltFAR	Street Seg.	BuiltFAR	Street Seg.	BuiltFAR	Street Seg.	BuiltFAR
s	5.3	x1	8.9	x2	7.6	x3	8.6
x4	8.2	x5	6.6	x6	6.6	x7	9.3
x8	7.2	X9	8.2	x10	8.2	x11	7.2
x12	7.0	x13	8.5	x14	8.8	x15	8.5
x16	9.6	x17	8.2	x18	7.6	x19	7.3
x20	7.6	x21	6.6	x22	9.6	x23	9.4
x24	7.2	x25	7.1	x26	8.6	x27	11.2
x28	7.2	e	7.7

,

Table 14. Noise density evaluation and visualization.

Number of Noise	Points in the Circle	s, x1~x28, e: Noise Density (311 Complaints Data from NYC Open Data): Intersections in a Zigzag Order Starting from the Lower Left in the Location on the Map (Table 4)
Street Seg.	Density. Noise	Street Seg.	Density. Noise	Street Seg.	Density. Noise	Street Seg.	Density. Noise
s	19	x1	29	x2	26	x3	10
x4	17	x5	5	x6	13	x7	36
x8	4	X9	31	x10	8	x11	6
x12	5	x13	28	x14	29	x15	10
x16	13	x17	2	x18	12	x19	17
x20	7	x21	2	x22	14	x23	28
x24	5	x25	2	x26	10	x27	11
x28	10	e	8

,

Table 15. Tree density evaluation and visualization.

Number of Trees	Points in the Circle	s, x1~x28, e: Tree Density (30 Points in Midtown Manhattan): Intersections in a Zigzag Order Starting from the Lower Left in the Location on the Map (Table 4)
Street Seg.	Density. Tree	Street Seg.	Density. Tree	Street Seg.	Density. Tree	Street Seg.	Density. Tree
s	69	x1	27	x2	21	x3	71
x4	19	x5	5	x6	31	x7	28
x8	1	X9	16	x10	61	x11	4
x12	0	x13	18	x14	18	x15	14
x16	19	x17	0	x18	19	x19	71
x20	1	x21	6	x22	24	x23	30
x24	5	x25	2	x26	67	x27	11
x28	23	e	56

,

Table 16. Visibility evaluation and visualization.

Number of Visibility	Visibility in the Circle	s, x1~x28, e: Visibility (30 Points in Midtown Manhattan): Intersections in a Zigzag Order Starting from the Lower Left in the Location on the Map (Table 4)
Street Seg.	Visibility. Area	Street Seg.	Visibility. Area	Street Seg.	Visibility. Area	Street Seg.	Visibility. Area
s	190	x1	119	x2	122	x3	115
x4	118	x5	167	x6	116	x7	115
x8	131	X9	125	x10	121	x11	115
x12	151	x13	119	x14	109	x15	112
x16	121	x17	120	x18	118	x19	113
x20	123	x21	126	x22	119	x23	133
x24	126	x25	118	x26	114	x27	217
x28	116	e	206

,

Table 17. Sky Exposure evaluation and visualization.

Number of Sky Ex.	Sky Ex. in the Circle	s, x1~x28, e: Sky Exposure (30 Points in Midtown Manhattan): Intersections in a Zigzag Order Starting from the Lower Left in the Location on the Map (Table 4)
Street Seg.	Sky Exposure	Street Seg.	Sky Exposure	Street Seg.	Sky Exposure	Street Seg.	Sky Exposure
s	52.83	x1	31.94	x2	26.63	x3	31.68
x4	31.83	x5	49.14	x6	30.56	x7	25.55
x8	35.86	X9	33.89	x10	27.93	x11	32.78
x12	40.73	x13	25.85	x14	30.02	x15	32.11
x16	25.90	x17	36.95	x18	38.10	x19	40.30
x20	34.06	x21	39.49	x22	29.03	x23	31.84
x24	35.74	x25	36.25	x26	28.59	x27	42.64
x28	36.11	e	44.07

and

Table 18. Landmark View evaluation and visualization.

Number of Landmarks	View in the Circle	s, x1~x28, e: Landmark View (Blue: Visible/Red: Invisible): Intersections in a Zigzag Order Starting from the Lower Left in the Location on the Map (Table 4)
Street Seg.	Num. Landmark	Street Seg.	Num. Landmark	Street Seg.	Num. Landmark	Street Seg.	Num. Landmark
s	1	x1	0	x2	0	x3	0
x4	1	x5	1	x6	3	x7	0
x8	1	X9	1	x10	1	x11	1
x12	1	x13	0	x14	1	x15	1
x16	0	x17	2	x18	1	x19	1
x20	3	x21	0	x22	0	x23	2
x24	0	x25	1	x26	0	x27	2
x28	2	e	3

show the results of collecting and evaluating the seven types of data across 30 points, visualized in a multidimensional format. Not only are the measured quantitative values represented, but the spatial density is also depicted in a 3D metaball format. Additionally, results from each point are presented in a continuous manner, visualized similarly to an animation.

2.4. Final Map Generation

Dijkstra Algorithm

Dijkstra’s Algorithm is an algorithm that solves the problem of finding the shortest path between two points. It is used in Google Maps and many other navigation systems, and it determines the shortest route by iteratively reviewing the available options from the start point to the end point.

Table 19. Final formula diagram: Dijkstra Algorithm.

Dijkstra’s Path Algorithm

(start point a; end point e) and

Table A2. Dijkstra Algorithm Python script.

Dijkstra’s Path Algorithm in Grasshopper

Def dijkstra(graph,src,dest,visited=[],distances={},predecessors={}): if src not in graph: raise TypeError(‘The root of the optimal path tree cannot be found’) if dest not in graph: raise TypeError(‘The target of the optimal path cannot be found’) if src == dest: # We build the optimal path and display it path=[] pred=dest while pred != None: path.append(pred) pred=predecessors.get(pred,None) print(‘optimal path: ‘+str(path)+” total score: “+str(distances[dest])) else: if not visited: distances[src]=0 for neighbor in graph[src]: if neighbor not in visited: new_distance = distances[src] + graph[src][neighbor] if new_distance < distances.get(neighbor,float(‘inf’)): distances[neighbor] = new_distance predecessors[neighbor] = src visited.append(src unvisited={} for k in graph: if k not in visited: unvisited[k] = distances.get(k,float(‘inf’)) x = min(unvisited, key = unvisited.get) dijkstra(graph,x,dest,visited,distances,predecessors) p1 = y1, p2 = y2, p3 = y3, p4 = y4, p5 = y5 p6 = y6, p7 = y7, p8 = y8, p9 = y9, p10 = y10 p11 = y11, p12 = y12, p13 = y13, p14 = y14, p15 = y15

p16 = y16, p17 = y17, p18 = y18, p19 = y19, p20 = y20 p21 = y21, p22 = y22, p23 = y23, p24 = y24, p25 = y25 p26 = y26, p27 = y27, p28 = y28, p29 = y29, p30 = y30 p31 = y31, p32 = y32, p33 = y33, p34 = y34, p35 = y35 p36 = y36, p37 = y37, p38 = y38, p39 = y39, p40 = y40 p41 = y41, p42 = y42, p43 = y43, p44 = y44, p45 = y45 p46 = y46, p47 = y47, p48 = y48, p49 = y49 if __name__ == “__main__”: #import sys;sys.argv = [‘‘, ‘Test.testName’] #unittest.main() graph = {‘s’: {‘x1’: p1, ‘x2’: p2}, ‘x1’: {‘x3’: p3, ‘x4’: p4, ‘s’: p1}, ‘x2’: {‘x4’: p5, ‘x5’: p6, ‘s’: p2}, ‘x3’: {‘x6’: p7, ‘x7’: p8, ‘x1’: p3}, ‘x4’: {‘x1’: p4, ‘x2’: p5, ‘x7’: p9, ‘x8’: p10}, ‘x5’: {‘x2’: p6, ‘x8’: p11, ‘x9’: p12}, ‘x6’: {‘x3’: p7, ‘x10’: p13, ‘x11’: p14}, ‘x7’: {‘x3’: p8, ‘x4’: p9, ‘x11’: p15, ‘x12’: p16}, ‘x8’: {‘x4’: p10, ‘x5’: p11, ‘x12’: p17, ‘x13’: p18}, ‘x9’: {‘x5’: p12, ‘x13’: p19, ‘x14’: p20}, ‘x10’: {‘x6’: p13, ‘x15’: p21, ‘x16’: p22}, ‘x11’: {‘x6’: p14, ‘x9’: p15, ‘x16’: p23, ‘x17’: p24}, ‘x12’: {‘x7’: p16, ‘x8’: p17, ‘x17’: p25, ‘x18’: p26}, ‘x13’: {‘x8’: p18, ‘x9’: p19, ‘x18’: p27, ‘x19’: p28}, ‘x14’: {‘x9’: p20, ‘x19’: p29}, ‘x15’: {‘x10’: p21, ‘x20’: p30}, ‘x16’: {‘x10’: p21, ‘x11’: p23, ‘x20’: p31, ‘x21’: p32}, ‘x17’: {‘x11’: p24, ‘x12’: p25, ‘x21’: p33, ‘x22’: p34}, ‘x18’: {‘x12’: p26, ‘x13’: p27, ‘x22’: p35, ‘x23’: p36}, ‘x19’: {‘x13’: p28, ‘x14’: p28, ‘x23’: p37}, ‘x20’: {‘x15’: p30, ‘x16’: p31, ‘x24’: p38}, ‘x21’: {‘x16’: p32, ‘x17’: p33, ‘x24’: p39, ‘x25’: p40}, ‘x22’: {‘x17’: p34, ‘x18’: p35, ‘x25’: p41, ‘x26’: p42}, ‘x23’: {‘x18’: p36, ‘x19’: p37, ‘x26’: p43}, ‘x24’: {‘x20’: p38, ‘x21’: p39, ‘x27’: p44}, ‘x25’: {‘x21’: p40, ‘x22’: p41, ‘x27’: p45, ‘x28’: p46}, ‘x26’: {‘x22’: p42, ‘x23’: p43, ‘x28’: p47}, ‘x27’: {‘x24’: p44, ‘x25’: p45, ‘t’: p48}, ‘x28’: {‘x25’: p46, ‘x26’: p47, ‘t’: p49}, ‘t’: {‘x27’: p48, ‘x28’: p49}} dijkstra(graph,‘s’,‘t’)

present diagrams explaining this process and the Python-based scripts used in the Grasshopper environment. In this study, we applied the Dijkstra path algorithm, a representative path-finding algorithm, to derive optimal routes reflecting user preferences. To do this,

Table 20. Nodes for Dijkstra Algorithm.

Nodes for Finding Optimal Path (p1~p49)

shows how 30 points at the target site are labeled (start point: s; end point: e; path point: x; and value between points: p). The intersections along the path from the start point s to the end point e are sequentially labeled x1 to x28 as selectable adjacent points. The measured values between the designated points are marked as P, and both points are assumed to be equidistant, with the average of the data from the two points representing the value.

Furthermore, the seven measured data types are reevaluated into four grades (with landmarks considering visibility from three surrounding towers being evaluated across three grades). The Dijkstra Algorithm is applied by reorganizing the data so that the first grade has the shortest path value p, following the ranking based on data size, from the highest to the lowest (Grades 1 to 4). To achieve this, the quartile method is used. Quartiles divide the data sample into four equal parts when sorted in ascending order and are defined as the first quartile (Q1), the second quartile (Q2), and the third quartile (Q3). This method is employed because data may cluster around specific values, making it essential to understand the spread and reorganize the data results accordingly. For example, for a sample dataset like −5, −2, 1, 2, 4, 5, 6, 8, 9, 15, 20, the median value Q2 is 5, with Q1 (the median of values smaller than Q2) being 1, and Q3 (the median of values larger than Q2) being 9. In this study, as shown in

Table 21. Grades and score range.

Grade 1 (Excellent)	Grade 2 (Good)	Grade 3 (Fair)	Grade 4 (Poor)
100~75%	75~50%	50~25%	0~25%
Score > Q3	Q3 ≥ Score > Q2	Q2 ≥ Score > Q1	Q1 ≥ Score

, the data are categorized as follows. The smallest data value up to the lower 25% (Q1) is classified as Grade 4, from Q1 to the middle 50% (Q2) is Grade 3, from Q2 to the upper 25% (Q3) is Grade 2, and from Q3 to the largest data value is Grade 1. To visually express the data deviation, a box plot method (a type of quartile representation) is used (

Table 22. Evaluated grades and score range.

Name of Data	Q1	Q2	Q3	Min./Max.	Grade 1 (G1)	Grade 2 (G2)	Grade 3 (G3)	Grade 4 (G4)
Sky Exposure	30.02384	33.34217	38.10313	25.55815/52.83286	G1 > 38.10313	38.10313 ≥ G2 G2 > 33.34217	33.34217 ≥ G3 G3 > 30.02384	30.02384 ≥ G4
Landmark View	0	1	1	0/3	G1 > 1	G2 = 1	G3 < 1	-
Visibility	116,102.4	119,981	126,939.4	109,256.2/217,190	G1 > 126,939.4	126,939.4 ≥ G2 G2 > 119,981	119,981 ≥ G3 G3 > 116,102.4	116,102.4 ≥ G4
Noise	6	10.5	19	2/36	G1 > 19	19 ≥ G2 G2 > 10.5	10.5 ≥ G3 G3 > 6	6 ≥ G4
Restaurant (Yelp)	5	8	12	1/17	G1 > 12	12 ≥ G2 G2 > 8	8 ≥ G3 G3 > 5	5 ≥ G4
Built-FAR	7.230833	7.967095	8.684286	5.30351/11.23344	G1 > 8.684286	8.684286 ≥ G2 G2 > 7.967095	7.967095 ≥ G3 G3 > 7.230833	7.230833 ≥ G4
Tree	5	19	30	0/71	G1 > 30	30 ≥ G2 G2 > 19	19 ≥ G3 G3 > 5	5 ≥ G4

, Figure 4). However, for the Landmark View category, due to the small range of data (0–3), it is classified into three grades. For the Noise category, event counts were classified in connection with the Comfort and Activity categories, where the Activity (high values) category is graded 1 to 4, and the Comfort category is defined in the reverse order. Additionally, decimal data values are rounded at the last decimal place.

3. Results

This study proposes a route guide map that explores the city based on three main characteristics—activity, comfort, and nature—allowing for personalized exploration according to user preferences, as shown in

Table 23. Final path.

Optimal Path: Activity, Nature, Comfort

,

Table 24. Final route: Activity.

Final Map—Preference: Activity

,

Table 25. Final route: Nature.

Final Map—Preference: Nature

and

Table 26. Final route: Comfort.

Final Map—Preference: Comfort

. Routes are guided based on the categorized data results from the density of these three factors, and the routes are displayed in Grasshopper as shown in Table 23. The node value p between each route point is defined as the average value of the two route points and is rounded to the nearest decimal place (

Table 27. Nodes and value: Activity.

Node	Value	Node	Value	Node	Value
p1	2.50	p2	2.17	p3	2.83
p4	2.67	p5	2.33	p6	2.67
p7	2.50	p8	2.67	p9	2.50
p10	2.50	p11	2.83	p12	2.83
p13	2.33	p14	2.17	p15	2.33
p16	2.67	p17	2.67	p18	2.67
p19	2.67	p20	2.83	p21	2.67
p22	2.33	p23	2.17	p24	2.67
p25	3.00	p26	2.17	p27	2.17
p28	2.33	p29	2.50	p30	2.67
p31	2.33	p32	2.67	p33	3.17
p34	3.17	p35	2.33	p36	2.17
p37	2.33	p38	2.83	p39	3.17
p40	2.50	p41	2.50	p42	2.83
p43	2.67	p44	3.00	p45	2.33
p46	2.33	p47	2.67	p48	2.67
p49	2.67

,

Table 28. Nodes and value: Nature.

Node	Value	Node	Value	Node	Value
p1	1.83	p2	1.83	p3	2.67
p4	2.83	p5	2.83	p6	2.33
p7	2.67	p8	3.00	p9	3.17
p10	2.67	p11	2.17	p12	2.17
p13	2.50	p14	3.17	p15	3.50
p16	3.17	p17	2.17	p18	2.83
p19	2.83	p20	3.00	p21	2.83
p22	2.67	p23	3.33	p24	3.17
p25	2.33	p26	2.33	p27	3.00
p28	2.67	p29	2.83	p30	3.00
p31	2.83	p32	2.33	p33	2.17
p34	2.83	p35	2.83	p36	2.17
p37	2.00	p38	2.67	p39	3.17
p40	2.33	p41	3.00	p42	3.00
p43	2.50	p44	2.17	p45	2.33
p46	2.67	p47	2.67	p48	1.33
p49	1.67

and

Table 29. Nodes and value: Comfort.

Node	Value	Node	Value	Node	Value
p1	2.83	p2	2.17	p3	3.17
p4	2.00	p5	1.33	p6	2.00
p7	3.00	p8	2.50	p9	1.33
p10	2.33	p11	3.00	p12	3.00
p13	2.50	p14	1.83	p15	1.33
p16	2.17	p17	3.00	p18	3.17
p19	3.17	p20	3.33	p21	2.17
p22	2.17	p23	1.50	p24	2.17
p25	2.83	p26	2.33	p27	2.50
p28	3.00	p29	3.17	p30	1.83
p31	1.83	p32	1.83	p33	2.50
p34	3.00	p35	2.50	p36	2.33
p37	2.83	p38	1.83	p39	1.83
p40	2.00	p41	2.50	p42	2.50
p43	2.33	p44	1.50	p45	1.67
p46	2.33	p47	2.33	p48	1.33
p49	2.00

). The data collected from the urban environment reflect the dynamic nature of city spaces and provide a foundation for adapting to changes up to date. Traditional function-oriented maps fail to account for individual preferences, remaining static and rigid. By contrast, the activity-centered route map allows users to explore and remember the ever-changing activities that take place in the city. The comfort-centered route map enables users to enjoy an attractive city by recognizing open views and landmark buildings while experiencing a noise-free, pleasant environment. Lastly, the nature-centered route map allows for mental refreshment through encounters with nature in the city. These personalized maps foster a more intimate relationship between the city and its users, ultimately contributing to the creation of more sustainable urban environments.

4. Discussion

Walking is the most fundamental mode of transportation for covering certain distances. It can be viewed as crucial from two perspectives. First, walking allows for free movement and provides high accessibility to surrounding facilities [50]. This creates opportunities for in-depth urban exploration, which can lead to urban revitalization. Second, walking incurs no cost and helps avoid issues such as traffic congestion, pollution, and noise that arise from vehicular transport. Therefore, providing appropriate walking paths to enhance walkability forms the foundation for creating environmentally friendly cities [51].

Numerous studies on pedestrian route selection have been conducted by traffic planners, focusing mainly on distance and time [24,25,26,27,28,29]. Meanwhile, some architectural and urban experts have continued to discuss how the built environment affects walking [52,53,54]. Key factors considered in these discussions are physical and aesthetic surroundings, with the environment being assessed based on safety [55], comfort [56], and interest [57].

This study characterizes the scope of analysis as areas perceptible from a pedestrian’s point of view, and the sequence of analysis was chosen based on the continuous nature of walking experiences. Furthermore, public data and location-based data collected over sufficient periods at the urban scale were utilized, ensuring both the reliability and accessibility of the data. By quantifying surrounding built environment data via computer analysis, this study offers a new alternative to the past survey-based research methods, which were time consuming and reliant on subjective judgment and individual experience.

In addition, this study employed 3D Rhino and Python-based Grasshopper as modeling tools to conduct data collection and evaluation tasks. These tools are highly advantageous for building 3D environments, essential for architectural and urban designers. Since they are based on the widely used Python programming language [58], they may offer greater opportunities for collaboration with a wider range of users compared to GPS tracking systems integrated with smartphones or GISs like ArcGIS, which are often used in recent research.

Lastly, this study includes a multidimensional data visualization process. This allows for interaction based on dimensions, provides 3D spatial information, and helps users effectively recognize urban-based data through a computer-based technique that expresses sequential experiences with continuous imagery [59,60]. Through detailed selections like 3D views, detailed information, and presentation methods, the data can be displayed more clearly, precisely, and effectively, increasing users’ visual engagement and sparking interest.

5. Conclusions

This study makes contributions to related fields in three main ways. First, it aids in understanding various characteristics of cities based on urban big data. By connecting public data and geo-location-based data with open-source platforms, this research provides a case study where projects can be applied without incurring additional costs. In other words, it offers clues for uncovering hidden aspects of today’s urban spaces, which are increasingly complex, through quantitative analysis. Second, it demonstrates the potential for expanding the use of data across the fields of urban planning, architecture, and transportation through the development of commercial applications or related product releases. This opens up opportunities to link urban architecture with new business fields and explore possibilities for commercialization. Third, by offering new ways to explore cities, this study contributes to urban revitalization and provides inspiration for creating walkable, pedestrian-friendly cities. This can lead to a positive feedback loop, increasing the utilization of surrounding commercial, public, and cultural spaces, thus enhancing the overall value of walking.

In summary, this study leverages urban data to evaluate pedestrian environments and categorizes the findings into three distinct types, proposing a user-centered route map that offers personalized urban experiences. The final map includes 3D route information, detailed numerical data from measurements, and multidimensional visualizations of data results at various points. This approach aims to bring about changes to the traditional methods of suggesting routes, which were based on uniform criteria, by exploring the potential for new data-driven designs that prioritize visual diversity based on quantitative analysis. Furthermore, this study suggests new methods for collecting and utilizing data, evaluating and visualizing pedestrian environments, establishing the relationship between urban spaces and individual user experiences, and applying relevant algorithms to spatial projects important areas for future similar projects. It is hoped that this work will provide insights into urban activation and contribute to sustainable urban planning. However, the appropriateness of limiting alternatives to three categories should be further examined, and data reliability is paramount to ensure accurate analysis. Therefore, continuous attention should be given to refining data collection methods, collection periods, the accuracy of surrounding urban modeling, and appropriate data visualization techniques. Finally, future research could expand by conducting detailed comparative analyses of how people living in different cities perceive their environments and how they are influenced by specific environments.