Please note that JRE (Java Runtime Environment) is required to run TetrisAI, A Typist's Calendar, ImageSearch, and AudioSearch.
A neural network model to perform sentiment classification of free-text movie reviews (positive or negative).
Combining latest neural network architectures and techniques such as
convolutional neural network (CNN),
recurrent neural network long short-term memory (RNN-LSTM),
and attention pooling.
Automated Scoring of Chatbot Responses in Conversational Dialogue.
Developed machine learning models
(traditional and neural network models) to score the quality of
chatbot responses in conversational dialogue setting.
Web application to aggregate and visualize social media activities.
Facebook, Twitter, and Instagram OAuth was implemented.
User's data are freshly pulled and not stored to ensure privacy.
Visuocial includes sentiment analysis and topic detector feature too.
A simple and light web application to recognize music titles being played on any YouTube Videos.
Simply paste the YouTube URL and watch the video while waiting for the result.
Android App to show latest PSI, real-time weather, and forecast info in Singapore.
Updated in real-time with Singapore NEA database (data.gov.sg).
Developed a Stochastic Part-of-Speech Tagger (Hidden Markov Model, supervised learning).
Kneser-Ney smoothing and interpolation were implemented to handle zero count.
Viterbi Algorithm was implemented to find the optimal sequence of the most probable tags.
AI to play Tetris and obtain best score possible.
Implemented 3 steps look-ahead and Genetic Algorithm for better performance.
Developed "A Typist' Calendar" Application using Java and Eclipse IDE.
Features:
1. Application to schedule our everyday activities/to-do list.
2. Standalone application, compatible with Microsoft Windows or Mac OS X.
3. Runs with or without internet connection.
4. Flexible and clear input/command format.
5. Synchronization feature with Google Calendar (requires internet connection).
Application for Image retrieval.
Search using an image and retrieve relevant images from an image database.
Features include Colour Histogram, SIFT, Visual Concept Recognition, and Text Description.
Multimedia content analysis and search.
Thesis title:
Automated Diagnosis of Acute Appendicitis Based on Clinical Notes (link)
Collaborated with National University Hospital (NUH) to develop a patented novel neural network model for
predicting acute appendicitis using free-text emergency department (ED) notes written by doctors.
Focus area:
• Artificial Intelligence
• Information Retrieval
Awards:
• IEEE Singapore Computer Society Book Prize (2016) (link)
-- Best Bachelor of Computing Honours Project in AY 2015/2016
• Certificate of Distinction in Information Retrieval (2016) (link)
• Certificate of Distinction in Artificial Intelligence (2015) (link)
Yuwono, S. K., Ng, H. T., & Ngiam, K. Y. (2019). Learning from the Experience of Doctors: Automated Diagnosis of Appendicitis Based on Clinical Notes. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL 2019), 18th BioNLP Workshop. (link)
Yuwono, S. K., Biao, W., & D’Haro, L. F. (2018). Automated Scoring of Chatbot Responses in Conversational Dialogue. In Proceedings of International Workshop on Spoken Dialog System Technology (IWSDS 2018), WOCHAT. (link)
Yuwono, S. K., Ng, H. T., & Ngiam, K. Y. (2017). System and Method for Computerized Diagnosis Based on Clinical Notes. Singapore Patent No. 10201709877S.
Yuwono, S. K., Ng, H. T., & Ngiam, K. Y. (2016). Automated Anonymization as Spelling Variant Detection. In Proceedings of the 26th International Conference on Computational Linguistics (COLING 2016), Clinical NLP Workshop. (link)
Past 4 years of working experience including internship and part-time positions.
Working on Android platform and ecosystem.
Responsible for building models and applying machine learning and NLP in healthcare domain: anonymization, data cleaning, model building and tuning, and clinical trial/validation set-up.
Developed novel neural network model to predict multiple diagnosis and length of stay using Emergency Department clinician's free-text notes
Software Engineer in the Core Experience (CXE) Team in Microsoft Office.
Involved in developing internal tools to enhance Microsoft Office development.
Developed tools to process, visualize, and aid developers to locate memory leaks, and detect critical thread blocking (i.e. UI) in multi-threaded environment.
Software developer in Media and Entertainment (M&E) department, Maya® Animation.
Embarked on Maya® Time Editor project.
Involved in developing new features and bug fixes of various scale and type including:
1. Back-end functionalities
2. Graphical User Interface
3. Maya Scripting Language (MEL)
Involved in organizing and preparing Singapore Maya User Group (SMAUG),
a workshop for Maya® Animation Customer to experience first-hand new features that are yet to be released.
Software Engineer and Data Scientist. Specializing in Machine Learning, Neural Networks, and Natural Language Processing (NLP).
Passionate about any kind of software development and research in computing-related field.
Freelance web and mobile app developer.
Feel free to contact me for any queries.
On the side, selling affordable and fully customizable musical instrument cables:
• 1/4 inch guitar cable
• XLR microphone cable
• and more
Please click here for more details.