- About
- Message from the Chair
- History
- Facilities
- News
- Events
- Info Sci Colloquium
- Advancing Responsible AI with Human-Centered Evaluation
- Bowers Distinguished Speaker Series - Julie E. Cohen, Georgetown University Law Center
- From Agents to Optimization: User Interface Understanding and Generation
- The Language of Creation: How Generative AI Challenges Intuitions—and Offers New Possibilities
- IS Engaged
- Graduation Info
- Info Sci Colloquium
- Contact Us
- Courses
- Research
- Computational Social Science
- Critical Data Studies
- Data Science
- Economics and Information
- Education Technology
- Ethics, Law and Policy
- Human-Computer Interaction
- Human-Robot Interaction
- Incentives and Computation
- Infrastructure Studies
- Interface Design and Ubiquitous Computing
- Natural Language Processing
- Network Science
- Social Computing and Computer-supported Cooperative Work
- Technology and Equity
- People
- Career
- Undergraduate
- Info Sci Majors
- BA - Information Science (College of Arts & Sciences)
- BS - Information Science (CALS)
- BS - Information Science, Systems, and Technology
- Studying Abroad
- MPS Early Credit Option
- Independent Research
- CPT Procedures
- Student Associations
- Undergraduate Minor in Info Sci
- Our Students and Alumni
- Graduation Info
- Contact Us
- Info Sci Majors
- Masters
- PHD
- Prospective PhD Students
- Admissions
- Degree Requirements and Curriculum
- Grad Student Orgs
- For Current PhDs
- Diversity and Inclusion
- Our Students and Alumni
- Graduation Info
- Program Contacts and Student Advising
Title: Data Vision - Learning to See through Algorithmic Abstraction
Abstract: Algorithmic data analysis has come to enable new ways of producing and validating knowledge. Algorithms are now integral to many contemporary knowledge practices, especially ones that rely on the analysis of large-scale manually intractable datasets. A large part of the appeal of data analytic algorithms is the seemingly abstract and mechanical nature of their application. We are often told that a specific algorithm can work on multiple datasets as long as the datasets are modeled in particular ways. This is of course a great source of algorithmic strength: if the hallmark of real-world empirics is its richness and unpredictability, the hallmark of data analytics is its ability to see and engage the world via abstract categorization and actionable manipulation of the world. However, data analytic algorithms often require a great deal of iterative, situated, and discretionary work to make them work with datasets. Being a data analyst, then, is the ability to see, organize, and analyze the world as, with, and through data, algorithms, and numbers, while simultaneously mastering forms of discretionary judgment in the ways in which worlds and tools are put together. This involves knowing why, how, and when to apply and improvise around established algorithmic rules and routines.
In my ongoing research, I demonstrate and study the situated nature of data analytics, showing how moments of application function as sites of routinization, discretion, and ‘professional vision.’ I argue that data analysis is an iterative, reflexive, and ongoing process, requiring data analysts to straddle the competing demands of methodological abstraction and empirical contingency. In this presentation, I focus on one particular aspect of my work: how do algorithms come to be seen as abstract and mechanical applications of well-defined rules? Drawing on research in HCI and the social sciences, and empirical data from a four-month long participant observation of a machine learning classroom, in this talk I will show how in a data analytic classroom an algorithm's application gets conceptualized, demonstrated, and understood as a rule-like sequence of abstract steps. This research is part of my collaborative work with Steve Jackson that was recently submitted as a CHI paper.