<aside> 🎨 Hello! I'm Yugin Tan, a Human-Computer Interaction researcher based in Singapore. I am a software engineer by training, but my passion lies in understanding how to best design digital systems that benefit society.
I am doing my Ph.D in HCI at the National University of Singapore (NUS) under Asst. Prof. Lee Yi-Chieh.
</aside>
<aside> 🌈 About Me: I am always thinking about how technology, society, and people interact in different ways. Consequently, I’m always looking for new projects to get involved with!
Outside of the lab, I love gaming, making music 🎹🎤, F1 🏁🇳🇱🇪🇸 and karting, and cats 🐱
</aside>
This lovely cat is, sadly, not mine - but I recently adopted two of my own!
👷🚧👷🚧👷🚧 Website under construction! 🚧👷🚧👷🚧👷
Please contact me for more information on any of the projects below: @slyffleclaw (Telegram), [email protected]
Social service agencies and practitioners help a wide range of people in society, such as those with financial, familial, and mental or emotional difficulties. Since October 2023, when two local agencies reached out to us to explore how generative AI can enhance the efficacy and efficiency of their operations, I have been exploring all the different ways we can use cutting-edge technology to help practitioners in this field do what they do best - helping the most vulnerable members of our society.
The projects we are working on include:
Beyond our work with social work agencies, I am also curious in the word-level design of chatbots. Using local context as an example, Singlish (i.e. Singaporean English) is universally spoken and understood by Singaporeans, and is associated with a certain degree of friendliness and comfort. But native use of Singlish is nuanced and measured, and naive attempts to instruct LLMs to use it results in uncanny and bizarre dialogue. Should chatbots speak Singlish, and to what degree? Can its careful and conscientious use improve user outcomes, especially in areas like emotional support and counselling where building rapport with the user is of paramount importance?
🎉 ICA ‘25 Best Student Paper! 🎉 (non-archival) Co-1st author; congrats to co-author Song Tianqi!
We are encountering intelligent digital agents increasingly often in daily life, such as with social bots on Twitter/X or chatbots on customer service websites. Can a group of these bots, through perceived social pressure, exert collective social influence on us in the same way as a group of human users would? We find that the same viewpoints expressed by larger numbers of digital agents indeed better persuade users to change their opinions, and continue to investigate this phenomenon across different contexts.
🎉 CHI ‘25 Best Paper! 🎉 2nd author; congrats to 1st author Dr. Zhu Zicheng!
Digital agents interact with us in many human-like ways, including by asking for assistance to complete particular tasks. Can helping an AI agent create psychological well-being benefits, similar to how helping our fellow humans often improves our emotional state? We find that assisting an AI chatbot with a simple conversational task results in increases in participant well-being, which are mediated by established psychological factors like perceived autonomy and competence.