Announcing Our Summer 2023 Undergraduate Research Fellows
Each year, the Berkeley Center for New Media pairs undergraduates with a graduate student mentor, offering them the chance to complete real, graduate level research while at Cal. We are thrilled to announce this semester's Fellows.
Akira Ono
Akira is an Ethnic and Media Studies double major, who has previously taught traditional art and worked in UC Berkeley's Queer Alliance Resource Center. Akira has a deep passion for their intersectional identities as a queer nonbinary Asian American and as a transgender Japanese person looks forward to contributing to piecing together Japan's queer history to increase visibility and uplift the voices of current trans people. Akira will be working with Lani Alden on Trans Theatrical Voices in Japan.
Lani project is working to preserve Japanese trans theatrical voices. Akira will assist Lani with metadata entry from a collection of materials from the national Diet online library, creating an important database of trans history in Japan.
Isbael Li
Isbael Li is a computer science major, pursuing the Jacobs Design and BCNM New Media certificate. Isabel has served as an intern at the Berkeley Institute of Design and at Apple, where she helped build a platform used by internal AI/ML research teams for developing machine learning. Previously, Isabel worked with Elnaz Tafrihi on InsightXR, to allow inclusive participation in AR applications. Isabel will be working with Gowri Swamy on Identifying Misinformation research.
The dissemination of AI-generated visuals has become a powerful weapon in spreading misinformation, yet the policies guiding its deployment are yet to be securely established, reported, and executed. This research aims to take a closer look into how users identify misinformation, their thoughts and perceptions towards AI-generated content, and which stakeholders are responsible for the truthful dissemination of information. The research questions posed act with the main objective of addressing the forward-looking cybersecurity policies that must be created to ensure safe, secure, and trustworthy online environments for today’s adolescents. This research will focus on the policies we can design to mitigate the impacts of mis/disinformation created by AI-generated synthetic media by taking a cross-stakeholder approach to understand its current and future implications on society, media, journalism, and technology.
Isabel will play a role in data analysis processes. Data collected (qualitative research interviews and results from situational mapping research task) will be analyzed using qualitative interview coding, affinity mapping, and thematic analysis.
Nika States
Nika States is a candidate in Geography, with an interest in sonic geographies, public infrastructure, behavioral science, and field recording.Nika has worked as a performing and recording artist and has at times incorporated field recordings into releases. Nika will be working with Adrian Montufar on Sonic Representations.
Sonic Representations is a sound art project about walking at night using sonic representations of street lights. Adrian is combining field recordings of walks on the UC Berkeley main campus with a spatialized sonification of the street lights and other nighttime illumination encountered. Nighttime lighting is one of the many elements of urban infrastructure that influences the paths we take while walking. It is connected to safety, both in real, practical terms, and in our imaginations—in our collective, contested understandings of what happens in the dark. First, the field recording documents the moment of the walk through space. Second, the sonification makes audible something that was already present in our experience but that wasn’t captured in the recording.
Reica Ramirez
Reica is a Political Science and Media Studies major, with a concentration in Media Law & Policy. Reica has served as a public relations coordinator for the Pilipinx Academic Student Services and a communications assistant for the Goldman School of Public Policy. Reica will be working with Sophia Perez on Mapping Pacific Islander Filmmakers.
The project collects data on Pacific Islander filmmakers and film locations and maps the information. Pacific Islanders are vastly underrepresented in American media and the project seeks to make it easier to dive into the rich, if lesser-known world of Pacific Islander film for people of all ethnicities, but particularly in support of Pacific Islander storytellers and audiences. This map will also ideally by a helpful tool for Pacific Islander storytelling grant-makers like Pacific Islanders in Communication, Nia Tero, and the Pasifika Entertainment Advancement Komiti; these organizations can use this map to ensure that storytellers from under-represented islands are given priority when applying for funding. Any data-based discoveries about historical absences of funding for Pacific Islanders can also be used in federal and/or international grant applications by filmmakers and other cultural organizations attempting to even the playing field.
Wish Wang
Wish Wang is a Computer Science and Cognitive Science double major. Wish is currently a research assistant at the Whitney Lab for Perception and Action, where Wish is working on the project “Interaction between Context and Emotion Perception.” Wish contributed to the code using PyTorch which trains an artificial intelligence-based model that keeps track of the emotion tracking of humans. Wish has basic data analysis skills and is interested in learning more about unsupervising learning algorithms for databases, and has had experience training deep-learning models in the past. Wish will be working with Xinwei Zhuang on Architecture Databases.
Xinwei's study proposes constructing a domain-knowledge-curated architecture database, equipped with a calculated similarity matrix to unveil the interconnection between projects. The curated database is designed for architects and architectural students to search for relevant architectural precedents. Researchers first collect projects with essential information about an architecture project, including structured data, such as architect, location, program, and scale, and unstructured data, such as photos, drawings, and, if possible, recorded video experience. They then clip the text and image data together to train a deep-learning model. The metadata sets can then be clustered and represented in a reduced dimension with unsupervised learning algorithms such as spectral clustering. After dimension reduction, when querying for a specific type or style of a building, one can explore relevant architectural projects sampled from their neighborhood from the latent space.