Keri Mallari

I am a PhD student at UW HCDE advised by Dr. Gary Hsieh. My research interests are in the field of human-computer interaction and social computing. I received my undergrad degree from CUNY Lehman College with a BS in CS and a BA in Math support by the Macaulay Honors Program.

I am a 2021 Twitch Research Fellow.

Research

Right now, I'm thinking about how to support content creators through the lens of online feedback exchange. Below are some of my pas and present projects.

Analytics Needs of Video Game Streamers

Conducted an interview study to describe the ecology of analytics tools utilized, define streamers' areas of analytics needs, and identify opportunities to support streamers and their communities.
CHI 2021

Exploring Expertise And Algorithmic Tuning on Human-AI Teams

Ran a user study that examined users' interaction with three simulated algorithmic models. Then, we explored what underlying factors significantly impacted the overall team performance of users' decision making with the recommendations of an AI-assistant.

Content Distribution in English Wikipedia

Collected and analyzed 600k article pages from English Wikipedia to uunderstand how they reflect a distribution of topical content. Then, explored content distribution of events based on decadal temporal analysis using Wikipedia's link structure.

HCI Science Communication on Twitter

Collected and analyzed tweets from HCI researchers to understand their science communication practices and opportunities to support them. Designed and developed an web app to visualize science communication efforts.

Experience

I've spent my summers learning from really cool places and really smart folks below:

Research Science Intern, Twitch

Explored and analyze trends and behaviors in the messages of small content creators. Designed a prototype reimagining chat analytics of content creators. Ran interviews and user studies to develop insights on prototype and develop analytics.
2021

Research Intern, Microsoft Research

Utilized computer vision using OCR and entity recognition to extract information and develop insights to support personal informationmanagement. Developed a full stack web application using nodejs and azure cognitive services.
2020

Research Intern, Microsoft Research

Investigated human-AI complementarity in high-stakes decision making scenario. Designed differentways to incorporate machine recommendations to leverage human-AI complementarity.
2019

Research Collaborator, Microsoft Research

Designed and ran a study to examine the impact of different presentations of racial information onhuman judgment in the context of recidivism. Developed a full stack web application to gather data on MTurk and analyzed results to study significant differences across conditions.
2018
CHI 2020

Data Science Summer School, Microsoft Research NYC

Analyzed student trajectory in the NYC public school system by calculating student performance basedon test results, and then tracking individual student performance over the years. Developed predictive models for dropout rates of students in the system and acceptance rates of students in the NYC highschool application process.
2017
MIT CODE 2017

Teaching

I've also TA'd some cool classes listed below:

HCDE 538: Designing for Behavior Change

Students are introduced to existing behavior change theories, frameworks, and research to gain an understanding of why and how behavior changes. Utilizing these insights, students practice theory-driven design to nudge positive behavior change. They analyze current behavior change applications and utilize existing resources to guide their design process.
Spring 2021

HCDE 539: Physical Prototyping

Students in this course design and prototype interactive prototype systems and environments using physical computing tools (a combination hardware and software). This quarter they have been exploring a range of applications using the Circuit Playground Express micro-controller (in Winter 2021), and an Adafruit M4 Feather (in Fall 2021).
Winter 2021, Fall 2021

HCDE 451: UX Prototyping

Students in this course have been exploring various user experience prototyping techniques such as model making, laser cutting, 3D printing, sewing, video, and interactive wire framing.
Fall 2020

DRG in Analyzing and Supporting Twitter-based Science Communication

In this project, we will be designing a system that identifies what types of people have likely read an individual tweet (e.g., using techniques from NLP and beyond), and aggregates that information to help us summarize the current state of science communication in HCI.
Spring 2019

Testimonials

"She is a smurf student" - Will, after catching Keri play Diner Dash in class

"Those are not pancakes" - Minsuk, after watching Keri make breakfast on stream