😶 
ጿ ኈ ቼ ዽ ጿ ኈቼ ዽ ጿ ኈ
Education
- Major : Department of Computer Science and Information Engineering
- Thesis : Temporal Graph Embedding incorporating with Memory Mechanism and k-hop Selection
- Advisor : Chao-Lung Yang & Shan-Hsiang Shen
- GPA : 4.02/4.3 ( Department Website , Transcript , Graduate Thesis )
- Major : Department of Money and Banking
- GPA : 3.82/4.3 ( Department Website , Transcript )
Experiment
- DevOps Engineer @ Open Source Innovation Group ( OSI Tech Blog , 凱西島的怪獸們 )
- Data Analyst @ SCSB Digital Banking Department - AI Lab
- Composed work instructions and operation guides
- Intern @ SCSB Digital Banking Department
- Designed and implemented a sentiment analysis system for scsb to monitor and interpret public opinion trends
- Master student @ SiMS Lab, NTUST ( SiMS Lab , Collaborative IT Blog by Graduate Lab Members )
- Enthusiast @ Knowledge Sharing ( IT Blog )
- You don't have to be great to start, but you have to start to be great
Competition & Award
- Hosted by taiwan academy of banking and finance
- Ranked as the 8th out of 187 teams
- Competition's Host Website , Slide , Group Photo , Participation Certificate
- Hosted by national kaohsiung university of science and technology - department of electrical engineering
- Ranked as the 1st out of an estimated 30 teams
- Competition's Host Website , Slide
Interest
- Building a python package is repetitive, so i want to automate tasks like setting up pre-commit hooks, documentation, and ci/cd.
- Create a template that includes pre-commit hooks, documentation, and ci/cd to simplify package setup.
- The template uses rye for dependencies, sphinx for documentation, docker for ci/cd, and github actions for automation.
- This template makes it easier to release packages, so i can focus on development instead of setup.
- As a software engineer, i explore open-source projects on github but often forget starred repositories.
- How to avoid forgetting starred repositories?
- Github actions schedule tasks, pygithub creates issues, and a state machine with pydantic manages reviews.
- The tool, still in progress, helps me review repositories regularly and effectively.
- A small project for converting the code snippet to the video.
- As a graduate student, i need to write my thesis in latex, but the original template is outdated.
- Refactored it for use on overleaf and local environment.
- Sought guidance on docker image creation specifically for xelatex and reviewed related templates.
- Shared the template on overleaf, a docker image, and tutorials for beginners.
- A python package for automatically download the image from the markdown file and save it to the local folder.
- Trying to use evasion attack to attack the cnn-based maleware detection model.
- Need to add noise to the maleware sample to make the model misclassify the sample.
- The dataset is collected by ourselves (using cuckoo sandbox and virusshare).
- In this challenge, i had written 30 articles about network science and graph neural network.
- Including the basic concepts of network science, the application of network science, like graph neural network and graph database.
- Also including the latest research papers about graph neural network.
- Using dynamic web scraping packages (selenium, playwright etc.) to crawl social network data.
- The data will be stored in mongodb and can be used for further analysis.
- These crawlers are only used for personal skill improvement.
- Applying network science field to build a keywords recommendation system.
- Finding identical products from item's name is a quite challenging task.
- Using regular-expression and ckip-transformer to extract keywords from item's name.
- And using dimension-reduction and cluster algorithms to find identical products.
- An application for analyzing the skills required by the job market.
- The data is crawled from 104 job bank and stored in mongodb.
- And the data will be processed by ckip-transformer.
- Finally, using apyori algorithm to find the association rules between skills.
- These crawlers are only used for personal skill improvement.
- Using django to build a web application for analyzing financial public opinion.
- The data is crawled from anue and the data will be stored in mongodb.
- These crawlers are only used for personal skill improvement.
- These crawlers are only used for personal skill improvement
- These crawlers are only used for personal skill improvement
- A line-bot for getting the stock information.
-
2025-09-23 - Higgs-Audio Common Token Summary -
2025-09-22 - Docling + ibm-granite/granite-docling-258M -
2025-09-18 - Introduction to the Easy Approach to Requirements Syntax (EARS) -
2025-07-24 - Docker Model Runner Installation -
2025-07-21 - ngrok - static domain for free -
2025-07-10 - AWS encounters InvalidToken or InvalidClientTokenId -
2025-06-19 - Using AWS SAM to Create and Run Lambda Functions -
2025-06-13 - A toy example of scanning models -
2025-06-06 - First Experience with Codex -
2025-05-12 - OCI Artifact?