Chunxu Wang

Research Assistant | Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong, 06/2024 to 08/2024

• Collected Twitter posts in the United States about the COVID-19, using Twitter API and web scrapper
• Conducted sentimental analysis on the Twitter posts, aiming to find the sentimental trend of US public during pandemic
• Used Dynamic Sentiment Analysis method to analyze sentiment in a temporal way over time, resulting in better performance

Sentimental Analysis in Social Media (Twitter) | Research Assistant, supervised Prof. Siyuan Liu  Sept. 2021 - June. 2023 | Pennsylvania State University

• Collected Twitter posts in the United States about the COVID-19, using Twitter API and web scrapper
• Conducted sentimental analysis on the Twitter posts, aiming to find the sentimental trend of US public during pandemic
• Used Dynamic Sentiment Analysis method to analyze sentiment in a temporal way over time, resulting in better performance

Football Scoring & Teaching using Machine Learning | Researcher  Sept. 2022 - Dec. 2022 | Wuhan University

• Researched on modern information technology to create a modern education information environment, on infiltrating and applying collected information to football teaching, promoting the development of student’s teaching innovation
• Initial results showed that applying machine learning techniques to teaching has a far-reaching significance for football players’ post-match scoring, in both motivation level and grade evaluation

Natural language Processing | Researcher with Professor Wenpeng Yin Sept.2023 - Dec.2023 | Pennsylvania State University

• Used CNN and RNN to do the Sentiment Analysis on Movie Reviews, aimed at classifying the sentiment of sentences extracted from the Rotten Tomatoes dataset. The objective is to discern whether a given sentence expresses positive or negative sentiment.

The Impact of Negative Platform News on User Behavior, Secondary-author    Dec. 2021 | The Pennsylvania State University

• Study shows that negative news report not only has effect on company's business strategies, but also on its user behaviors. Data (100k+) sourced from Bilibili, where Difference-in-Differences measure and Rubin Causal Model were used to quantify user behavior, which was then imported into Hypothesis Testing on user metrics between control and treatment group.
• Suggestions to state-run news agency: pay more attention to the impact of negative news on customer-facing industries.
• Paper published as: He, Wang, Xu, & Fan. (n.d.). The Impact of Negative Platform News on Users Behavior. International Conference on Machine Learning, Big Data and Business Intelligence.