**5.1 Introduction**

*Sustainability in Urban Planning and Design*

**4. Challenges of participatory design**

that are present in user participation approaches.

be drawn from the raw user data.

expertise.

visualize in a proper way for future interpretation.

temporal and spatial patterns.

opportunities for making better sense of public space through technology. One of the key benefits of adopting a data-driven approach to urban analytics surveys is the ability to see a combination of datasets in context with each other and to detect

A new generation of researchers has been deriving evidence-based rules for urbanism, which benefits from user participation [6]. These rules replace outdated working assumptions that have created dysfunctional urban conditions. Recent methodologies in urban research validate human scale urbanism and collaborative approaches. In order to provide a better understanding of the contradictory approaches, we will list some of the main challenges of centralized urbanism. Moving beyond the form-oriented framework of centrally based urbanism, we should also refer to certain challenges that the participatory approach entails. The growing desire of involving participants in the process represents certain challenges that need to be addressed for successful decision-making [7]. Building user participation systems in response to the complexity requires a combination of data, which is fit for use and decision support tools. We list some of the key barriers

1. Complex data user inputs. User data inputs are usually complicated data types. For example, natural language text, descriptions, sketches are a challenge for computers to interpret and also for researchers to translate them into a binary or measurable form. This type of data is also difficult to store, categorize, and

2.The translation of miscellaneous forms of data input is a labor intense, manual analysis and might result in potentially obscuring part of knowledge that can

3.Ensuring that the user understands the request and is able to provide useful feedback. Abstract requests could result in user distraction, which can complicate the feedback data previously described in the first point. Moreover, researchers will need to consider that the user input should not rely heavily on the users technical skills and prior knowledge of the tools, as this would limit the target user group to a very small pool of people, which would have the

4.Citizens are often a resource of small-scale ideas that could improve the livability of their immediate environment. However, it is hard for local people to coordinate and produce visualized results that they could communicate with the authorities. Even so, such proposals are likely to be discarded as they do not represent the stakeholders' benefits and moreover, large-scale developers make

it impossible for citizens to have any influence in urban development.

Based on the above, opening a channel for sharing knowledge and opinions is not necessarily sufficient for building a system that takes the most advantage of user input. The objective is to achieve a balanced relationship between extensive information and clarity, in order to ensure that all the data and their interconnections are handled to their entirety. We need to build human-computer interaction in a way that it facilitates user orientation and comprehension of the framework, defines the

**152**

The first example is a mapping process of the gentrification and displacement rate and livability levels in the neighborhoods of Oakland in the San Francisco Bay Area. Before analyzing the methodology of the example, we should first understand the notions of neighborhood and gentrification as addressed in this chapter, which will provide clarity regarding the reasoning behind the example methodology.

The neighborhood is often understood as the physical building block of the city for both social and political organization [8] and thus combines physical and nonphysical characteristics. Early scholars have described neighborhoods as defined, closed ecosystems, characterized only by their physical elements, such as size, density, demographics, etc. that would get disrupted by external factors, such as new residents. Moreover, neighborhood change has been regarded as a natural process of population relocation and competition for space, until a state of equilibrium could be reestablished. Based on these ideas, neighborhoods were presented as a deterministic model and categorized based on simplified criteria such as their residents' financial status, etc. However, neighborhoods are not introverted, autonomous clusters, and the mechanisms of neighborhood change do not rely on exclusively external factors. According to Jacobs [1], nowadays, people identify a neighborhood by a landmark in the city because it has become intimate from daily use or encounter. The key that creates the notion of a neighborhood is diversity and identity. She argues that people tend to avoid visiting places that do not represent any variation either in function or esthetics [1]. Although the modern way of living has urged people to be more mobile than previously, people tend to pay attention to district that surrounds their home if it meets the certain criteria that fit their lifestyle. The stability of a neighborhood relies on its capacity to absorb opportunities and sustain its diverse character. In this paper, the term neighborhood can be described as an instance of organized complexity [1].

The notion of gentrification can be described as one category of neighborhood change and is broadly defined as the process of improving and renovating previously deteriorated neighborhoods by the middle or upper class, often by displacing low-income families and small businesses. The first documented use of the term "gentrification" [9] describes the influx of a "gentry" in lower income neighborhoods. Owens identifies nine different types of neighborhoods that are experiencing upgrading: minority urban neighborhoods, affluent neighborhoods, diverse urban neighborhoods, no population neighborhoods, new white suburbs, upper middle-class white suburbs, booming suburbs, and Hispanic enclave neighborhoods [10]. Gentrification does not only rely on a singular cause, as it may emerge when more than one condition is present. It is a complicated process that does not rely on binary and linear explanations. Early studies identified two main categories that cause gentrification: private capital investment for profit-seeking and people flow that refers to individual lifestyle preferences [11]. Gentrification does not necessarily result in negative effects, as it can also operate as a tool for revitalization. When revitalization occurs from existing residents, who seek to improve their neighborhood conditions, the result can be constructive in enforcing the neighborhood stability. This condition is called incumbent upgrading or "unslumming' as Jacobs [1] defines it. However, when revitalization causes the displacement of current residents and a decline in neighborhood diversity, then neighborhoods gradually become segregated by income, due in part to macrolevel increases in income inequality as well as decline of job opportunities. Hence, neighborhood stability is compromised because the opportunities have been narrowed down to a very limited range of financial status and lifestyle. Displacement, however, is identified as the biggest negative impact of concern resulting from neighborhood revitalization and gentrification. Displacement occurs when any household is forced to move from its residence, usually because of eviction and unaffordable rent increase [12]. However, tracking unwilling displacement can be challenging to categorize, as researchers have faced limitations regarding data availability and data comprehension.

In this case study, we carefully selected and analyzed the various, specific data sets that relate to gentrification and are associated with livability, from authoritative census data categories, such as income, crime, education level, employment rate, urban infrastructure, etc. to more ephemeral and subjective data classifications related to human perception and user input. In order to go beyond the conventions in understanding the dynamics that drive socioeconomic phenomena and construct lived space, we attempted to implement methods that although they are considered disassociated with urban analytics, they offer a strong potential in contributing to this study as it will be analyzed in detail in the following paragraphs.

This case study involves three methods of data accumulation and analysis; the first method is a preliminary census data classification of key GIS data sets that are available from the government and other certified public resources. The second method uses data resources that derive from open data platforms (data that is freely accessible), such as Google API, Google Places, and collective, open-data platforms where users post all kinds of requests (sell and buy, real estate, etc.), such as "craigslist.org", while the third uses human perception and subjectivity as a qualitative source of data that can unveil qualities that could not appear otherwise and enrich the outcome with a diverse layer of data. This enrichment leads to a more informed decision-making and a more qualitative image of the city that reflects subjective aspects of urban planning [13].

Although the methods differ significantly in the types and source of data being used, it is important to mention that each perspective provides a different lens through which to view transition toward more or less livable and gentrified environments. The data sets collected from the three methods operate at different scales, some at urban scale for the entire San Francisco Bay Area, some at neighborhood scale, and some at street level scale. Each method presents certain advantages and altogether provide a calibrated understanding of the multiple grains of constructed space through top-down and bottom-up methods, as well as to offer a tool of visualizing dynamical characteristics of the urban environment. For example, using a human-based perspective alone may lead us to commit to something, which is entirely subjective, by ignoring holistic factors that emerge at aggregate levels and vice versa. The census data analysis provides an overview of the context over a significant time span (2000–2012) and helps us understand major socioeconomic shifts that affect tenure, which then affects the local market and the standards of living in the area in terms of public infrastructure. The open data analysis depicts the ephemeral layer of relationships that take place in the urban environment, which is impossible to be described by authoritative data; however, it is more relevant to the actual conditions, revealing virtual changes and dynamics for the near future. The third method enriches the process with user personal feedback about ranking the environment of a neighborhood as it currently stands.

**155**

*Human-Centered Approaches in Urban Analytics and Placemaking*

Across all three studies, the data has been visualized based on a few basic rules.

The initial method, which defines the preliminary step of the research, refers to the use of GIS census data analytics. This method aims to depict the urban areas that have undergone changes on authoritative parameters that are associated with the phenomenon of gentrification, such as tenure status, median household income, land value, and employment rate. Moreover, this method queries the livability levels based on parameters related to public infrastructure and the urban quality, such as pedestrian network continuity and status, transportation, walkability, and car dependency street trees and parks, schools, education points, medical and religious spaces. The objective of these data sets integration is to assess whether the built environment is evolving toward an equal state of services and other oppor-

The "Geographic Information System", or GIS, tools enable the accelerated gathering of data sets of multiple categories. As this method focuses solely on census data and basic population characteristics, the data sets that are useful to the execution of the survey are population, median household income, level of educa-

Initially, the survey began with a high-level analysis of the San Francisco Bay Area natural morphology and broad population characteristics; therefore, the first data sets that were visualized were green areas, wetlands, urban areas, total population, and land value (**Figure 1**). Before analyzing more thoroughly the city of San Francisco, we collected some data related to the homeless population community in the city as it is a very apparent phenomenon that appears to be getting aggravated. As the core of the research is focused on depicting and tracing the dynamics of gentrification, we believe that information associated with the homeless population community and then compared with the census data regarding tenure status and

The second stage of the survey focused solely in the county of San Francisco, as it was considered a suitable context to trace the main changes in demographic characteristics. The data sets that were visualized at this stage of the survey were the range of household income, range of home value, owner-occupied housing, vacant lots, and the ratio of unemployed population against the total population (**Figure 2**). Green areas in the San Francisco Bay Area that are accessible to the public, such as parks, plazas, etc, were excluded from the calculation as they do not relate to the

Although the previously described survey did reveal information on the transition of some neighborhoods in San Francisco, at the next stage of the research, it became apparent that the most applicable scale for census data display would be that of the entire San Francisco Bay Area. The reason for this is that census data has low spatial resolution and therefore refers to large-scale surveys. Hence, the data was recollected for the San Francisco Bay. The census data collected consists of data sets that range

targeted data sets, and they would have affected the results of the survey.

Changes of degree in a factor are displayed with a gradient of the same color, changes of type are displayed with different colors, and the general vocabulary of visual styles is communicated with dots, lines, and areas [14]. The tools used for the data visualization are "Microsoft Excel" for calculation of delta, median, and average values, "Grasshopper" for processing data input in ".json," ".csv," or ".shp" file formats, "Rhinoceros" for processing the output data from "Grasshopper," "Processing" as a geo-located three-dimensional virtual space, where multiple data sets can be displayed and overlaid at the same time, in order to assess their relation-

*DOI: http://dx.doi.org/10.5772/intechopen.89675*

ship and "Adobe Suite" for final printed output.

**5.2 Method 1: preliminary census data analysis**

tion, transport network, and infrastructure.

land value could provide useful insight (**Figure 1**).

tunities or in favor of certain socioeconomic groups over others.

*Human-Centered Approaches in Urban Analytics and Placemaking DOI: http://dx.doi.org/10.5772/intechopen.89675*

*Sustainability in Urban Planning and Design*

aspects of urban planning [13].

neighborhood conditions, the result can be constructive in enforcing the neighborhood stability. This condition is called incumbent upgrading or "unslumming' as Jacobs [1] defines it. However, when revitalization causes the displacement of current residents and a decline in neighborhood diversity, then neighborhoods gradually become segregated by income, due in part to macrolevel increases in income inequality as well as decline of job opportunities. Hence, neighborhood stability is compromised because the opportunities have been narrowed down to a very limited range of financial status and lifestyle. Displacement, however, is identified as the biggest negative impact of concern resulting from neighborhood revitalization and gentrification. Displacement occurs when any household is forced to move from its residence, usually because of eviction and unaffordable rent increase [12]. However, tracking unwilling displacement can be challenging to categorize, as researchers have faced limitations regarding data availability and data comprehension.

In this case study, we carefully selected and analyzed the various, specific data sets that relate to gentrification and are associated with livability, from authoritative census data categories, such as income, crime, education level, employment rate, urban infrastructure, etc. to more ephemeral and subjective data classifications related to human perception and user input. In order to go beyond the conventions in understanding the dynamics that drive socioeconomic phenomena and construct lived space, we attempted to implement methods that although they are considered disassociated with urban analytics, they offer a strong potential in contributing to

This case study involves three methods of data accumulation and analysis; the first method is a preliminary census data classification of key GIS data sets that are available from the government and other certified public resources. The second method uses data resources that derive from open data platforms (data that is freely accessible), such as Google API, Google Places, and collective, open-data platforms where users post all kinds of requests (sell and buy, real estate, etc.), such as "craigslist.org", while the third uses human perception and subjectivity as a qualitative source of data that can unveil qualities that could not appear otherwise and enrich the outcome with a diverse layer of data. This enrichment leads to a more informed decision-making and a more qualitative image of the city that reflects subjective

Although the methods differ significantly in the types and source of data being used, it is important to mention that each perspective provides a different lens through which to view transition toward more or less livable and gentrified environments. The data sets collected from the three methods operate at different scales, some at urban scale for the entire San Francisco Bay Area, some at neighborhood scale, and some at street level scale. Each method presents certain advantages and altogether provide a calibrated understanding of the multiple grains of constructed space through top-down and bottom-up methods, as well as to offer a tool of visualizing dynamical characteristics of the urban environment. For example, using a human-based perspective alone may lead us to commit to something, which is entirely subjective, by ignoring holistic factors that emerge at aggregate levels and vice versa. The census data analysis provides an overview of the context over a significant time span (2000–2012) and helps us understand major socioeconomic shifts that affect tenure, which then affects the local market and the standards of living in the area in terms of public infrastructure. The open data analysis depicts the ephemeral layer of relationships that take place in the urban environment, which is impossible to be described by authoritative data; however, it is more relevant to the actual conditions, revealing virtual changes and dynamics for the near future. The third method enriches the process with user personal feedback about

ranking the environment of a neighborhood as it currently stands.

this study as it will be analyzed in detail in the following paragraphs.

**154**

Across all three studies, the data has been visualized based on a few basic rules. Changes of degree in a factor are displayed with a gradient of the same color, changes of type are displayed with different colors, and the general vocabulary of visual styles is communicated with dots, lines, and areas [14]. The tools used for the data visualization are "Microsoft Excel" for calculation of delta, median, and average values, "Grasshopper" for processing data input in ".json," ".csv," or ".shp" file formats, "Rhinoceros" for processing the output data from "Grasshopper," "Processing" as a geo-located three-dimensional virtual space, where multiple data sets can be displayed and overlaid at the same time, in order to assess their relationship and "Adobe Suite" for final printed output.

#### **5.2 Method 1: preliminary census data analysis**

The initial method, which defines the preliminary step of the research, refers to the use of GIS census data analytics. This method aims to depict the urban areas that have undergone changes on authoritative parameters that are associated with the phenomenon of gentrification, such as tenure status, median household income, land value, and employment rate. Moreover, this method queries the livability levels based on parameters related to public infrastructure and the urban quality, such as pedestrian network continuity and status, transportation, walkability, and car dependency street trees and parks, schools, education points, medical and religious spaces. The objective of these data sets integration is to assess whether the built environment is evolving toward an equal state of services and other opportunities or in favor of certain socioeconomic groups over others.

The "Geographic Information System", or GIS, tools enable the accelerated gathering of data sets of multiple categories. As this method focuses solely on census data and basic population characteristics, the data sets that are useful to the execution of the survey are population, median household income, level of education, transport network, and infrastructure.

Initially, the survey began with a high-level analysis of the San Francisco Bay Area natural morphology and broad population characteristics; therefore, the first data sets that were visualized were green areas, wetlands, urban areas, total population, and land value (**Figure 1**). Before analyzing more thoroughly the city of San Francisco, we collected some data related to the homeless population community in the city as it is a very apparent phenomenon that appears to be getting aggravated. As the core of the research is focused on depicting and tracing the dynamics of gentrification, we believe that information associated with the homeless population community and then compared with the census data regarding tenure status and land value could provide useful insight (**Figure 1**).

The second stage of the survey focused solely in the county of San Francisco, as it was considered a suitable context to trace the main changes in demographic characteristics. The data sets that were visualized at this stage of the survey were the range of household income, range of home value, owner-occupied housing, vacant lots, and the ratio of unemployed population against the total population (**Figure 2**). Green areas in the San Francisco Bay Area that are accessible to the public, such as parks, plazas, etc, were excluded from the calculation as they do not relate to the targeted data sets, and they would have affected the results of the survey.

Although the previously described survey did reveal information on the transition of some neighborhoods in San Francisco, at the next stage of the research, it became apparent that the most applicable scale for census data display would be that of the entire San Francisco Bay Area. The reason for this is that census data has low spatial resolution and therefore refers to large-scale surveys. Hence, the data was recollected for the San Francisco Bay. The census data collected consists of data sets that range

**Figure 1.**

*Geo-located 3D space in the software processing. Natural elements and census data for San Francisco Bay Area. Public resources for homeless population for the city of San Francisco (figure was created by the author) [20, 21, 22].*

from 2000 to 2012 and is related to tenure status, median household income, median home value, and employment rate. For every data set of the above, we calculated the delta value (amount of change or difference) between the years 2000 and 2012 and remapped the values to a numerical range between 0 and 1, which corresponded to a gray scale ranges from white (255, 255, 255) to black (0, 0, 0). White color represents no change, whereas black color represents the highest amount of change. The delta value was plotted in the context of San Francisco Bay Area, and the result is four maps, each for one data set. The four maps, which derived from the process described above, represent the amount of change in tenure, median household income, median home value, and employment rate were weighted and integrated into a single map that represents the amount of change of all four data sets (**Figure 3**).

In addition to the data sets related to tenure status, we included the census data of artists' employment rate as it is considered a key indicator of the early stages of a gentrification process. Surveys in the field of urban renovation have established

**157**

**Figure 2.**

*population (figure was created by the author) [18].*

*Human-Centered Approaches in Urban Analytics and Placemaking*

the artists community as an agent of urban gentrification, for the reason that low-income artists tend to revalorize unproductive spaces because they are affordable and, as a result, increase the attractiveness of the neighborhood. Artists make the first move into post-industrial, post-welfare neighborhoods, and soon they attract the hipster movement before, eventually, being displaced by them and their new middle-class neighbors. Both participate in the cycle of exploring, developing

*Census data for the city of San Francisco, total population, income, vacant housing, home value, unemployed* 

*DOI: http://dx.doi.org/10.5772/intechopen.89675*
