My research interests focus on data policy and ethics through a critical analysis of artificial intelligence, surveillance technology, social media, patents, and the implications of civic rights and democracy through the data and systems that construct the datafied citizen.
Marija Dalbello and Catherine McGowan. “Memory Narrations as a Source for Historical Ethnography and the Sensorial-Affective Experience of Migration.” In Challenges and Solutions in Ethnographic Research: Ethnography with a Twist. Eds. Tuuli Lähdesmäki, Eerika Koskinen-Koivisto, Viktorija L.A. Čeginskas, pp. 161-184. New York: Routledge, 2020. View
Paris, B., Reynolds, R., & McGowan, C. (2021, April 22). Platforms Like Canvas Play Fast and Loose With Students' Data. The Nation. https://www.thenation.com/article/society/canvas-surveillance. View
ABSTRACT
Platform labor is a digital, capitalist infrastructure that leverages and exploits digitally performed human labor to increase profit margins. Within a platform labor model such as Amazon Mechanical Turk (MTurk), power asymmetries are essential to render human workers invisible and devalue their skills. Workers are become invisible through the use of Application Programming Interfaces (APIs) and algorithmic management functions. Human workflow is managed by algorithmic functions, and the product of human computational labor is collected through APIs; the use of digital technology as an intermediary between the human and her work renders her invisible through the elimination of human-to-human interaction. In this paper, I will explore the relationship between platform labor and invisible work, specifically within the context of MTurk and propose research to legitimize digital labor and transform invisible work into visible work. First, I will present the essential functions and characteristics of MTurk. Next, I will review existing literature to provide an overview of invisible work, present models of invisible work in platform labor models, and describe how human workers are rendered invisible specifically within MTurk. Finally, I will propose methods to explore invisible labor and the phenomenon of heteromation through the theoretical lens of political economy, coupled with data activist methods within critical informatics theory. I will propose these theoretical approaches to address the gaps in literature and present unanswered questions to address for future research.
ABSTRACT
In this paper, I will conduct a literature review in order to assess predictive policing and crime analysis tools that are utilized in criminal justice. Predictive policing technologies and big data are utilized to identify individuals or geographical regions where criminal activity could potentially occur through the use of big data, machine learning, and algorithms. Predictive policing technologies aid in identifying patterns expeditiously and can assist law enforcement to identify where to focus limited resources effectively. However, the use of predictive policing technologies and big data poses many ethical and political concerns such as: reinforcing racial biases; the inaccurate automation of police officer knowledge work; and the potential for the negative impact of due process for citizens through the influence of criminal justice laymen (data analysts). As such, my preliminary research questions are as follows: 1. How are algorithms and big data utilized to perform risk analysis or predictive policing? 2. How do algorithms and the use of big data impact the decision making process in policing? 3.How do predictive policing technologies impact the knowledge work of police? In 2011, Time Magazine declared a predictive policing tool one of the “50 best inventions of the year” (Grossman et al, 2011). Traditionally, the role of law enforcement is reactive to crime—a crime is committed and law enforcement responds to the reported event. Predictive policing tools leverage historical crime data (time, location, nature of crime) to conduct statistical analysis and make predictions for where, when, why, or by whom the next potential crime could be committed. Law enforcement agencies will use statistical crime prediction reports to inform decisions such as: where and what time officer patrol routes should be scheduled and which individuals should be flagged as a risk to society and be surveilled. However, criminological datasets are not collections of naturally occurring science; the definition of crime is socially constructed and evolves over time. Furthermore, it is important to recognize the role of race inequality in law enforcement. According to the Federal Bureau of Prisons statistics, 37% of incarcerated inmates are black and 32.2% of inmates are Hispanic (Federal Bureau of Prisons, 2019). Accordingly, it is imperative to consider how culture and society shape the meaning of crime and the stereotypical criminal suspects that are ingrained in criminological datasets.
ABSTRACT
The Internet is a modern tool for humans to share bad information quickly. Humans presently interact within digital social worlds on a daily basis, arguably on an hourly basis. Within these digital social worlds, individuals can create information, consume information, and share information. All types of information are easily recorded in momentary flashes, often constrained by the limitations of character count, recorded video time-length, the frame size for a photograph, or the visibility of the total content in one glance, without the need to scroll to see more. When an individual interacts within any one of the various digital social worlds available, there is an implication that humans have evolved (or, perhaps, devolved) to a space where we attempt to iterate the biggest messages on the smallest scales. The vastness, or power, of a message is quantifiable, predicated by common digital measures, viz.: Likes, Shares, Comments, and Retweets. In other words, information is now available, most commonly, in bite-sized packages, widely referred to as memes. While many researchers dedicate time to identifying information behavior, the types of information that are sought, and what that means for the development of knowledge, I am proposing that knowledge is not sought after or acquired consciously when interacting in a digital social world. Rather, the intent is to interact socially. In this paper, I will explore why the modern, information-seeking human will consume bad information, digest it, and then regurgitate it for others to consume. Before I can begin to unravel why an individual would share bad information, I will define and present an example of what bad information is. I will meditate on observations of modern human information behavior relative to individual epistemological processing, trust, and how information has the option of validity through the exploration of bad information.
ABSTRACT
How can the gap between a social digital world and a digital library be bridged? In this paper, I will propose a digital library that offers Targeted Information to users to bridge the gap between the digital social worlds of users and the information stored within the databases of a digital library. The proposed methods will employ data mining of a user’s browsing history to identify subjects and key terms that are current and relevant to the active and present interests of the user. Once the key terms are collected from the user’s browsing history, the website of a digital library will auto-generate recommended titles for the user to explore. This proposal is an attempt to connect a user’s Internet social activity with the content available within a digital library.
ABSTRACT
Cyber-bullying is frequently noted as a contributing factor for teen suicide. My question was: why? Why does cyber-bullying have such an incredibly profound psychological effect on teens? What is happening on a social media application such as Facebook that leads to deep psychological scarring and depression and cause a teen to choose suicide as the solution? I had to believe that there must be something more to a status update than merely self-expression. There must be deeper meaning to the status update than the surface level meanings of a selfie, a photo of a delicious meal, and the activity of performing for an audience that the term self-expression implies. It is through my questions and research that I discovered the concept of self-writing and how it relates to the Facebook status update.
ABSTRACT
What happens when we become accustomed to an algorithm? Social networking sites curate content for their users through the use of algorithms. Users feed the algorithm data that will determine how it will filter and return content for consumption. This is achieved through a combination of the user creating a user profile and the system logging the habits of the user. However, the algorithm is designed to search, filter, and retrieve for the user without providing any visibility to the process of the options available and the choices that were made—the process is opaque. So, what happens when the algorithm makes a choice that is different from what the user would have made? The purpose of this paper is to explore how the algorithms that are designed to curate content for users of social networking websites create regimes of ignorance. The opaqueness of (the human designed) algorithms creates a system that strips users (humans) of the ability to rationalize and make informed choices, and, arguably, denies the user agency in their own reality. I will expand on how algorithms demonstrate the existence of systemic ignorance by presenting cases where a user would have made a different choice than the algorithm. The phenomenon of systematized ignorance strips individuals of their autonomy to make choices that are informed, to rationalize and construct knowledge in conjunction with experience. Thus, ignorance is established and the irrational man is created.
ABSTRACT
What is digital folklore? How does folklore present itself in the digital age? What are examples of digitally born folklore? I started this paper with these presumably simple questions and discovered complex, indefinite answers. The focus of this paper centers on the premise of Internet memes as examples of digitally born folklore, due to their commonplace abundance in digital worlds, their proliferate mode of production and transmission that exceeds the expectations of viral entities, and their clear structural characteristics that can be defined as genres and employed as taxonomic means for further interpretive research. Memes are digitally born artifacts that are produced, transmitted, and housed within Internet Websites and the pages of Social Networking Sites. The nature of Websites and Social Networking Site pages are temporary. Content can remain indefinitely, content can be edited, and entire websites can be removed permanently, destroying all historical record of its contents. Due to the ephemeral nature of websites, it is important to capture and archive content to preserve the record of cultural exchange and values for historical documentation and to perform research. In this paper, I will draw parallels to folkloric methods of analysis to the observable genres of memes to demonstrate the potential of research for memes. I will also provide an overview of what memes are, the current work that defines meme genres, the difference between viral digital objects and memes, and the cultural exchange and participation in the creation and transmission of memes. Finally, I will review current issues with web-archiving and web-scraping methods through a demonstration of the archival issues I discovered through the American Folklife Center’s archive of memes, and the issues this presents for limiting adequate research.
Design and archive store communication directives for all Visual Merchandising, Graphics, and Fixture installations for a chain of 900+ stores in the US, Canada, and Mexico. Create and manage Graphics and Fixture Budget; annual fund is $7.5mil. Manage Creative Production and Store Communications teams to design and execute four quarterly seasonal packages and monthly packages for store rollouts. Develop merchandising fixtures for new products and departments. Revamped the New Store ordering process to customize kits by store, with a savings of $5-7k for each of the (100) New Store projects executed annually—a savings of at least $500k. Designed and developed installation directive format and vehicle, significantly increasing successful execution chain-wide by consistently capturing clear methods of communication and accurate detail. Designed a database for running queries to identify SKU discrepancies in a merchandising Plan-o-gram.