About
About Xinmin Chu
I work on the problem of making fragmented data answerable for itself — deciding when two records describe the same thing, combining views of data that disagree, and keeping the resulting structure honest about its evidence.
Research path
I am a PhD candidate in the Computational Science program at the University of Massachusetts Boston, where I also work as a research assistant. Before that I completed a master's degree in statistics and a PhD in economics. The route through economics and statistics into computation was not a detour: entity resolution sits exactly where those fields meet — measurement problems, inference under noise, and algorithms that have to hold up at scale.
My current research centers on entity resolution and deduplication: multiview representation, record linkage and blocking, similarity graphs and clustering, threshold calibration, and evaluation integrity. The public artifact of this program iserbot, an open-source R toolkit for end-to-end entity resolution. Earlier published work includes a systematic review of cross-cultural information systems research: Chu, X., Luo, X. R., & Chen, Y. (2019),Information & Management, 56(3), 403–417.
Teaching and building
I have taught across economics, statistics, numerical methods, and machine learning — fields that students often meet as separate subjects but that keep answering each other's questions. The same interdisciplinary habit drives the software I build:ERBOT as a research instrument,KnowledgeTree as a private beta on evidence-aware knowledge structures, and DreamReality as an experimental prototype for structured dream narratives. Each carries an explicit maturity label, because a claim about software should be as careful as a claim in a paper.
Beyond the lab
Research is one way of paying attention. I also observe animals, institutions, cities, landscapes, and rhythm—subjects that rarely fit into a single discipline but often reveal how systems actually behave. Writing on these subjects — including poems and fragments — is collected in Notes.