This is a website for a ACM CHI 2019 workshop on rapidly emerging topic of Computational Modeling in HCI that aims to address the challenges of increasing complexity of human behaviors we are able to track and collect today. The goal of this workshop is to reconcile two seemingly competing approaches to computational modeling: theoretical modeling, which seeks to explain behaviors vs. algorithmic modeling, which seeks to predict behaviors. The workshop will address: 1) convergence of the two approaches at the intersection of HCI, 2) updates to theoretical and methodological foundations, 3) bringing disparate modeling communities to CHI, and 4) sharing datasets, code, and best practices. This workshop seeks to re-establish Computational Modeling as a theoretical foundation for work in Human-Computer Interaction (HCI) to model the human accurately across domains and support design, optimization, and evaluation of user interfaces to solve a variety of human-centered problems.
Keynote
Everything Old is New Again: What Can We Learn from the Past?
This workshop’s description on the CHI Accepted Workshop’s page says “This workshop will start a discussion about a set of guidelines necessary to establish Computational Modeling as a theoretical foundation for work in Human-Computer Interaction (HCI) to model the human accurately across domains and support design, optimization, and evaluation of user interfaces…” I posit that computational cognitive modeling began as the first theoretical foundation for HCI (indeed Card, Moran and Newell’s 1983 The Psychology of Human-Computer Interaction coined the name of the field), so an important question in this discussion is “What can we learn from the past?” Having been a researcher, an educator, and a practitioner in this discipline for over 35 years, I bring an enthusiasm for the new and a wealth of questions and cautionary tales from the old.
Bonnie E. John has been involved with computational cognitive modelling since 1983 when, as a mechanical engineer, she began PhD work in Psychology to “beat psychology into a form that engineers can use.” Working with Allen Newell and Stu Card, Bonnie continued the work introduced in their seminal book The Psychology of Human-Computer Interaction, extending GOMS to a version that allowed perception, cognition and motor processes to work in parallel (CPM-GOMS), to more effectively model highly skilled human behaviour. As a faculty at Carnegie Mellon University, she did research in cognitive modelling across domains and predictive measures. She applied computational cognitive modelling in industry to evaluate proposed systems for efficiency before they are built or purchased. Bonnie taught various versions of GOMS in university courses, CHI and other conferences, private companies and government agencies, for two decades. She led a decade-long research program nicknamed “cognitive crash dummies” to bring the power of psychologically-valid computational models to UX practitioners requiring no knowledge of either psychology or programming. This research program produced a practitioner-oriented tool, CogTool, with full documentation and tutorial materials, and half a dozen research versions that demonstrated different predictive metrics and types of use, integrated with UX design tools and Software Engineering practices, identified new challenges, and proposed solutions for those challenges. Bonnie currently works as a UX Designer at Bloomberg L.P.
Important Dates
March 1, 2019 (11:59 GMT) – submission deadline (EasyChair).
March 7, 2019 – submission decisions.
May 5, 2019 – workshop.
Workshop Agenda
09:00 – Welcome and Introduction
09:20 – Keynote Speaker – Bonnie John (followed by Q&A)
10:20 – Mid-Morning Break
10:30 – Research Speed Dating (quick introduction to each position paper)
11:30 – Brainstorm Key Areas for Computational Modeling in HCI
12:20 – Lunch Break
13:30 – Breakout Groups
15:20 – Mid-afternoon Break
15:45 – Report Back from Breakout Groups
16:30 – Brainstorm Future Direction
17:30 – Workshop Concludes
Computational Modeling in Human-Computer Interaction 