HANG GU · 顾航
I study how models learn from different kinds of data — and how we can evaluate them with care. I bring a frontend engineer’s perspective to research, data systems, and teaching.
Lecturer & Teaching Section Head
江苏财经职业技术学院
Different signals.
Shared understanding.
MULTIMODAL LEARNING · THOUGHTFUL EVALUATIONHow can visual representations contribute when labeled examples are scarce?
01 / RESEARCH
Learning across modalities. Understanding what evaluation reveals — and what it can hide.
A manuscript on adapting frozen vision-language features, combining modalities at the token level, and examining data artifacts in low-resource fake news detection.
Investigating the sensitivity of validation in news classification when labeled data are limited.
Work on a vision-language approach to multimodal behavior prediction.
Master’s research using R and Bayesian networks to study dental deformities prediction among adolescents, at Universiti Malaya.
02 / ENGINEERING
A selection of public code and applied work, connecting research questions with practical systems.
Multimodal fake-news detection with CLIP features and co-attention. Explore the implementation and setup on GitHub.
External datasets and model assets required; setup documented in the repository.
Explore repositoryA small, inspectable path from raw data to a useful interface: a frontend dashboard, a Python / SQLite pipeline, and notes on frontend and Spark concepts.
Synthetic data · Educational example
Open the learning labAn interactive student-affairs workflow sample, with role views, ownership transitions, version checks and explicit retry contracts.
Static showcase · Synthetic interaction data · Selected code
Code, architecture & checksSelected project experience. Source code is not publicly available.
Applying document understanding and interface design to contract management workflows.
Exploring document retrieval and retrieval-augmented generation for information access.
A component-based approach to organizing administrative interfaces and reusable page modules.
My frontend practice also includes multilingual interfaces, weak-network scenarios, dynamic-form state machines, and reusable admin components.
03 / THE NOTEBOOK
Practical notes and small examples that make frontend and data concepts easier to inspect, revisit, and teach.
Browse the collectionFRONTEND
Dashboard interactions, semantic markup, state, and accessible data presentation.
DATA ANALYSIS
Synthetic data, reproducible transformations, SQL aggregation, and data-quality checks.
BIG-DATA CONCEPTS
Notes on Spark, partitioning, and distributed processing, alongside a small local learning pipeline.
04 / BACKGROUND
I am Hang Gu (顾航), a computer science lecturer and teaching section head at 江苏财经职业技术学院. Before moving into teaching, I worked in frontend engineering and team leadership.
My master’s study in data science at Universiti Malaya deepened my interest in research. I am now interested in doctoral research on multimodal learning, careful evaluation, and useful data-driven systems.
2021.09 — 2023.04
Master of Computer Science · Data Science
2011.09 — 2015.06
Bachelor’s degree · Computer Science
2023.08 — PRESENT
Lecturer & Teaching Section Head
2019.03 — 2022.07
Frontend Lead
2018.03 — 2019.02
Frontend Engineer
2015.12 — 2017.02
Frontend Engineer
READING GUIDE
Explore the project, inspect the implementation, then review its checks and limitations. Follow a short path through the research code, engineering cases and learning experiments.
KEEP EXPLORING