Reflection on Transforming, Visualizing, and Mapping Project Data
It is important for me to transparently share some of the challenges I faced when completing both theoretical and practical challenges in this project, as you might be wondering about similar problems, or facing them as well. The challenge I faced was working with a specific part of the publicly available data on the riots of 1981 and 2011, which can be critiqued to subjectively communicate and impose the interests of its creators onto the audience engaging with the data via interactive dashboards and maps. The creators include teams of journalists and government employees, who often document quantitative data useful for various economic/political/legal systems' objectives, and de-prioritize quantitative data that can reflect the protesters' motivations. While I was aware of the critique, most of the data I could access and work with (on riot incidents and arrests) was of such nature, coming from The Guardian's journalistic and The Home Office's governmental data sources. Thus I found it conflicting to choose to represent such data via new mediums if the data of potentially opposing type is not considered and illustrated in this medium too. However, incorporating several kinds of data, including the one the significance of which we aim to challenge, can more transparently portray the complexity of historical events. To overcome this issue, I utilized existing visualization tools to analyze dominant data sources more critically, highlighting the incompleteness and lack of verified evidence behind the records. I also aimed to incorporate data that could challenge the significance of more dominant data sources and visualize it in a way that could persuade the users to see both sources as equally valid sides of comparison when they consider multiple perspectives of analyzing the historical event.
Additionally, the work I completed, including the challenges I faced, helped me to find interest in theories and applications of data maximalism, subjective mapping, deep mapping, and generous interfaces, as their functionalities can help portray historical events in a more complex, multi-dimensional way, communicating multiple and/or opposing perspectives of different groups of people. Though, when I applied such concepts within existing software (ArcGIS StoryMaps, Data Studio, Gephi/Palladio, BigQuery), I often encountered technical constraints that could limit my deliverables' effectiveness. As a solution, a combination of different Digital Humanities tools might be more effective for implementing such theoretical and applied concepts in a larger project, along with the user-facing project documentation raising the use of such concepts. It can help communicate the objectives behind different parts of the project, the benefits and limitations of its DH approaches, and its impact on the users.
Below I reference some of the key theoretical frameworks I aimed to apply in three projects, along with a description of how I applied them, and more specific challenges that I faced, as you might be interested to learn about them a bit more:
| Project | Theoretical Framework(s) | Applied Approach(es) | Challenge(s) |
| 1: Cleaning Data | Data Feminism (Principles 1 and 2: Examining and Challenging Power) | The project motivated to query and explore local authority data separately, to differentiate inferences on riot events in local versus national contexts; the approach was implemented to challenge the homogenous "criminal" narrative of the 2011 riots in the U.K. due to prior governmental power dynamics in how its narrative on the riots was constructed and reinforced over time. | (1) Absence of the official dataset documentation; (2) numerous misspellings and discrepancies in the names of local authorities; (2) dataset initially cleaned without considering how data variables should be defined and classified to be ingested in Google’s BigQuery and Data Studio. |
| 2: Visualizing Data | (1) Generous Interfaces (data explorability and browsability); (2) Critiquing data source as an “Objective Source of Truth.” | (1) The project incorporated search, filter features, and multiple approaches to view and analyze the data; (2) a tabular representation challenged the objectivity of data by flagging the data records which lack evidence. | (1) The other values in "Source" were not flagged in red as evidence-lacking - e.g., when a value such as "Source_Title" is lacking a link and needs to be differentiated from the link format values, as not verified; (2) Google Data Studio constraints on cross-filtering (the histogram displays only one bar record when cross-referenced, instead of all records sharing the exact date/location as the selected one). |
| 3: Mapping Data | Remapping (using new knowledge to challenge the prior knowledge reflected in maps by similarly incorporating it into the maps and representing both maps as valid) | The project used the same base map layer settings to represent the governmental and rioters’-driven data on the 2011 riots, representing the latter as an equally valid perspective on the events, which can be reflected similarly in mapping. | (1) Data imbalance (28 records in the governmental data source versus 14 data records in the interviewed data source); (2) StoryMaps constraint of not being able to place two ArcGIS maps for comparison in one visualization. |