I'm Sansiri Tarnpradab (āļŠāļąāļāļāđāļŠāļīāļĢāļī āļāļēāļĢāļāļĢāļ°āļāļąāļ) from Bangkok, Thailand. I recently graduated with a Ph.D. from Computer Science department at the University of Central Florida, under supervision of Prof. Kien Hua in the Data System Group (DSG). My research interests span the fields of Machine Learning (ML), Deep Learning, Natural Language Processing (NLP), Social Media Analytics (SMA), and Data Mining.
University of Central Florida, Orlando, FL
Syracuse University, Syracuse, NY
King Mongkutâs University of Technology Thonburi (KMUTT), Bangkok, Thailand
Comprehensively validated functionalities of implemented image reconstruction algorithms by writing test functions, examining behaviors, and documenting each test. I was also responsible for generating Monte Carlo simulated data from GATE simulator.
I conducted research with an emphasis on Deep Learning, Natural Language Processing, and Social Media Analytics. My dissertation centers on the application of recurrent neural networks to address online social media issues.
During Spring '15 - Fall '16, I joined the team of the full-stack developers to implement the NASA-funded research project "TableTop", a multimedia communication platform that supports synchronized sharing and manipulation of multimedia objects for the launch team. Check out our project here!
Applied medical imaging algorithms, implemented the FPGA pairing policy to estimate the number of random coincidences from Positron Emission Tomography (PET) scanner, and simulated data from GATE software.
Assisted and graded assignments, quizzes and exams. Occasionally, I also gave a lecture as a substitute lecturer. (Class of 173 students)
Assisted lab sessions, held recitations, and graded assignments, quizzes, exams.
Designed and developed animation of the Android application for kids with autism.
Conducted research and study on various Data Mining algorithms under guidance of Prof. Masaomi Kimura. (ããããĻã!)
Sansiri Tarnpradab, Kien A. Hua
In this research, we propose an ensemble neural architecture to detect rumor on Twitter. The architecture incorporates word attention and context from the author to enhance the classification performance. The experiment on the real-world Twitter dataset demonstrates promising results.
In ICDIM, 2018
Oral presentation
Sansiri Tarnpradab, Fei Liu, Kien A. Hua
In this work we make the first attempt to adapt the hierarchical attention networks for thread summarization. The model draws on the recent development of neural attention mechanisms to build sentence and thread representations and use them for summarization.
In FLAIRS, 2017
Oral presentation
Implementation of several capabilities that are essential for PET. For instance, a pairing policy to estimate random coincidences, and pre-processing of big data simulated from GATE opensource.
A multimodal conferencing system to support teams management and rapid decision-making tasks. The system provides the functionality of real-time synchronized sharing and manipulation of multimedia objects.
Learn moreAn Artificial Neural Network based model developed to predict the amount of future streamflow at the Reynolds Creek Experimental Watershed, Southwestern Idaho, USA.
Learn moreAnalysis of online customersâ reviews for Windows Surface vs iPad to determine whether the advertisement successfully communicated people of features and whether the customers were satisfied with them.
Learn moreStudy of C4.5 algorithm to determine if defects occur in the production line of Hard disk drives (HDD). These defects were keys to finding root causes and proper solutions to increase HDD yield.
Learn moreImplementation a secure and reliable file backup system by applying a Hybrid algorithm which combines two algorithms: AES-CBC and RSA, to encrypt and decrypt files. The program was also able to detect any unauthorized changes.
Learn moreSansiri Tarnpradab
Department of Computer Science, HEC 250
University of Central Florida
4000 Central Florida Blvd
Orlando, FL 32816
sansiri@knights.ucf.edu