Hi, I am Ken ...

• Seeking for Machine Learning and Data Science Summer Internship opportunity, 2019.

NSF funded Ph.D. Computer Science Student at North Carolina State University.

• Data Science and Machine Learning Researcher at the RAISE Lab (NCSU).


During the day I'm a Data Scientist and Machine Learning Researcher in the making. I am passionate in solving real-world problems empirically especially exploring the synergy of SE development with Computational Science practice and understanding social network of researchers and how that tie mediate to a student’s performance and future scholar attainment. My prime areas of interest are Machine Learning, Data Science, and Algorithms specifically in Text Analytics and Graph Mining.

Beside daily learning and doing rocket sciences, I enjoy running, cooking, playing badminton, poetry, and playing board games.

Education

North Carolina State University

PhD in Computer Science August 2016 - Present

Graduate Merit Fellowship ($10,000+) for GPA 3.5+

Appalachian State University

Bachelor of Science in Computational Mathematics 2011 - 2015

Minor in Computer Science

Graduated with Magna Cum Laude, 3.80/4.0. Top 5% of my graduated class.

Work

North Carolina State University

Graduate Teaching and Research Assistant January 2015 - Present

  • Guided by Dr Tim Menzies, a pioneer in automated software engineering, search based software engineering and data mining in software engineering
  • Current Project: (1)studying how the computational science field maintain and develop their software projects and (2)analyzing the sentiment and purposes of scientific citations of SE research papers
  • Statistical tests to measure correctness of estimation techniques. ANOVA, A12, Bootstrap and Cliffs-Delta are some of the methods adopted
  • Coordinate with the professor & other Teaching Assistants as a team to consolidate plans, structure the course, design tests, conduct review sessions, facilitate labs, and deliver the lesson

YouNet Corporation

R&D Data Science Intern May 2017 - August 2017

  • Researched and designed scalable data-driven algorithms from machine learning and neural network methodologies.
  • Analyzed the semantics of data, developed model, optimized performance, & deployed projects to solve business problems.
  • Devised and enhanced core products in the system: Facial Beauty Rating, Facebook User’s Age Prediction with Graph Clustering and Text Analytics, & Car Detection.

OverMountain Studios

Web Developer Intern May 2016 - August 2016

  • Designed, built, and maintained client’s application with AngularJS & Bootstrap front-end and a NodeJS backend.
  • Tested functionality of client websites, troubleshoot for issues and re-structured websites for scalability and usage.

University Housing at Appalachian State University

Head Resident Assistant January 2013 - August 2015

  • Organized building wide and cross building programs that fostered community for 500+ students
  • Expressed the voice of Resident Assistants and students on campus through composing policy and legislation proposals from Resident Assistant Council to University Housing Leadership
  • Directly advised and collaborated with the new Resident Assistant(s) and Chair of the residence hall council professionally and efficiently to help develop their leadership
  • Ensured the enforcement of legal and university policy through campus wide duty rounds to provide a safe and inclusive living environment for all members of the App State community.

Publications

Is One Hyperparameter Optimizer Enough? - Accepted for SWAN 18
Can You Explain That, Better? Comprehensible Text Analytics for SE Applications - Submitted for ASE 18

Projects

Facebook User’s Age Prediction

Mined and investigated unstructured 25k Facebook users’ posts through Vietnamese language processing with traditional machine learning methods, convolutional neural networks, & LSTM to predict the user’s age. Achieved accuracy = 79.6%.

Facial Beauty Rating

Developed a facial attractiveness rater based on the SCUT-FBP dataset, contains images of 500 Asian women and applying open-source OpenFace software to extract facial landmarks as features. With Pearson correlation of 0.86, convolutional neural networks outperform traditional machine learning method such as Random Forest (only achieved 0.64).

Sentiment Analysis

Built a recommendation system with Apache Spark and Alternative Least Square algorithm integrated users’ reviews sentiment analysis along with other attributes on the 4.5+ million reviews and 146k+ businesses Yelp dataset to suggest new businesses that are appropriate to the users’ interests.

Graph Embedding for Recommender System

Implemented DeepWalk as a graph-based technique to recommend the viewer-movie pair by evaluating and preference propagation algorithms (word2vec model) in heterogeneous information networks generated from user-item relationships.

Developer Triage Bot

Developed a bot on AWS that assigns & suggests suitable tasks for the developer along with recommending experienced developers in the team to help them if needed in real time through Slack interface and Github task tracking.