Student theses from TMU Graduate Institute of Data Science, College of Management win recognition by international conference and international journal
Source: College of Management
Published on 2021-06-21
Under the guidance of Professor Chun-Wei Tung (童俊維) and Associate Professor Yung-Chun Chang (張詠淳), students Che-Yu Chou (周哲宇), Yu-Ya Cheng (鄭宇雅), Yan-Ming Chen (陳彥銘), and Wen-Chao Yeh (葉文照) from the TMU Graduate Institute of Data Science, College of Management shared their research findings in the format of either an oral presentation in the 19th Asia Pacific Bioinformatics Conference (APBC 2021) or a published paper in Applied Science.
Ensemble learning for predicting ex vivo human placental barrier permeability
Under the guidance of thesis advisor Professor Chun-Wei Tung, student Che-Yu Chou from the Graduate Institute of Data Science presented his paper entitled “Ensemble learning for predicting ex vivo human placental barrier permeability” as the first author.
Chou has developed a computer machine learning model for predicting the penetration of chemical substances in the ex vivo placental barrier.
Given ethical debate about vivo clinical trials, as well as the concerns cover their cost- and time-efficiency, the use of in vivo experiments is expected to decline. However, this research used data from ex vivo experiments to develop a model that utilizes machine learning methods to achieve good research results. This learning model will greatly enhance new drug development, drug management, and environmental safety testing in chemical substance screening.
Valence and Arousal-Infused Bi-Directional LSTM for Sentiment Analysis of Government Social Media Management
Under the guidance of their thesis advisor Professor Yung-Chun Chang (張詠淳), students Yu-Ya Cheng (鄭宇雅), Yan-Ming Chen (陳彥銘) and Wen-Chao Yeh (葉文照) from the Graduate Institute of Data Science jointly published a paper entitled “Valence and Arousal-Infused Bi-Directional LSTM for Sentiment Analysis of Government Social Media Management” in the international academic journal, Applied Science, in January 2021. The team used a novel sentiment analysis method to analyze text sentiment with more detailed sentiment information.
To effectively analyze the massive amount of public opinion data generated by social media, the team developed a deep neural network model for sentiment analysis using bi-directional long short-term memory model that integrates emotional dimensions to analyze sentiments toward governmental management of social media.
Based on the research, the students also delineated ways to understand public opinions and improve the efficiency of social media operations. The results of their study can be used in future commercial applications such as customer service and product/brand management.
Did you know?
|Asia Pacific Bioinformatics Conference (APBC) is annual event of great importance to the international bioinformatics community. Participated by thirteen countries in 2021, the conference this year primarily centers on practical applications for information science, in medical topics such as genetics, pharmaceuticals, toxicology, and molecular compounds. For data scientists and research institutes, APBC is one of the flagship events that facilitate debates and sharing in topics around precision health and health big data.|