Portfolio
Conversations at work sometimes lead to some very uncomfortable situations. Smalltalk illustrates exactly how often such conversations turn into sexual harassment. By utilizing official EEOC data, Smalltalk shows the frequency of sexual harassment in the workplace for 20 different industries and demonstrates how much women and men are affected.
I worked on the data analysis, design, and implementation with the mentorship of Joshua Wilson. My overall goal for this project was to create an interactive visual representation of real-world sexual harassment data that would raise awareness of the issue by increasing its personal relevance to the user.
You can try Smalltalk on itch.io.
Building Connections
A Participatory Design Project with Muslim Refugee Women in Germany
Germany hosts a large number of refugees; many of them from predominantly Muslim countries. This has led to sometimes heated public debates about how to integrate those refugee communities into German society. The rhetoric of those debates often positions Muslim refugee women as helpless, oppressed victims rather than individuals with self-determination and agency. In this context, I wanted to examine how communication technology can help refugee women make social connections in their host community.
I worked on this participatory design project as part of a class on Information and Communication Technology for Global Development. I collected interview data from refugee families and a German volunteer located in Wandlitz, Brandenburg. Based on this, I developed a number of design ideas, and mock-ups that could aid refugee women build social connections. I then gathered feedback on these preliminary design ideas.
The main features of the proposed designs focus on connecting refugee women with each other and with German members of the community by exchanging personal, cultural, or emotional experiences. Additionally, there is a large focus on visuals in order to reduce language barriers that often hinder communication.
A more detailed design documentation page is currently in progress. Feel free to email me at rlietz@ucsc.edu for any questions about this project.
Sadboi is a game that requires the player to express the right emotion at the right time. It uses voice input and machine learning to progress the narrative based on the primary emotion in the player’s voice. This game explores whether incorporating a player's emotions can enrich and/or better facilitate the game experience.
Sadboi was a collaborative project between me and Marjorie Ann Cuerdo as a final project in CM 202 - Computational Media Research. My role in this project was to train a neural net to accurately detect emotions from a sound file and create an API that could be called from the game side. The full project, including a formal write-up and links to all created, publicly accessible Colab notebooks can be found on its GitHub page.
Smartwatch App for Mood Detection
The goal of this project was to develop an intelligent wearable mood detection system. We developed a Fitbit Ionic smart watch app that uses heart rate, step count, and manual labeling to capture and predict users' mood.
I worked on this project as an undergraduate researcher with a group of fellow research assistants in the Heracleia lab. My role was to help with system design and data collection. We presented this research at the PErvasive Technologies Related to Assistive Environments Conference (PETRA ’19) with the following publications:
Lietz et al. (2019). A wearable system for unobtrusive mood detection. https://doi.org/10.1145/3316782.3322743
Lietz et al. (2019). Survey of mood detection through various input modes. https://dl.acm.org/doi/10.1145/3316782.3321543
Harraghy et al. (2019). A review of wearable heart rate sensors in research. https://dl.acm.org/doi/10.1145/3316782.3321550
High/Low
High/Low is a mood journaling app for mobile devices. It allows for daily journal entries detailing the user's high, low, and overall mood. Keyword analysis and mood graphs help the user view mood trends over time, while the search and calendar features make it easy to look at individual entries.
This app was made for Alli K Design by Joshua Wilson and me. I contributed to the back-end data processing by writing a Rapid Automatic Keyword Extraction (RAKE) algorithm that is used in the search feature of the app. High/Low is available for purchase at: Google Play Store, Apple App Store, Amazon App Store
Some fun artsy projects
Party like it's 2020: Imagine the COVID-19 pandemic happened in the 80s. This is what it would’ve looked like. Humans all in their individual boxes, plenty of distance between them, and absolutely no fun. Created using lvllvl.
A Day in the Valley: I imagine this would be the start screen to a cute lil adventure game. An alien child gets lost while playing in the valley and needs to find its way back home without getting hurt by falling fruit. Created using ZX81 DOODL.
Snowy Chase: It’s a starry night somewhere high up north. Everything is peaceful. A herd of deer is passing through the village. The only disturbance in this peaceful night is a hungry wolf chasing its dinner. Created using lvllvl.
ERROR: Imagine the world was falling apart so much that not even error messages can keep it together. Created using ZX81 DOODL.
Confectionary Compress: Candy Crush but on an 80s console with Japanese art style. Fun fact: people who would’ve played this would look down on modern-day Candy Crush because they hate mobile games. Created using lvllvl.
Polygon Flower: Based on a photo I took and edited to create the Low Poly style. Created using Adobe Photoshop.