To keep its yearly tradition and to still give students an opportunity to present their research contributions under the impact of the pandemic, for the first time, the Division of Science and Technology (DST) transferred the 8th Science and Technology Poster Presentation online.

The event was held on 15 April, with 291 people having registered to watch the presentations and 144 people having voted for the posters.

Judges from inside and outside UIC, teachers and students watched the presentations through video conferencing software and voted through a Conference Registration System (CRS) designed by students from the Computer Science and Technology Programme.

This is the second time the DST Poster Presentation having used the CRS for voting

A total of 74 research projects from seven programmes got exhibited during this poster event. The topics included the psychology of preschool children, predicting stock movements using news reports, anti-diabetes research and Zhuhai's seagrass bed ecosystem.

The award-winning projects

In the opening remark, UIC Associate Vice President and Dean of DST, Prof Huang Huaxiong, explained that one of the biggest challenges for UIC during this difficult time has been how to minimise the impact of the outbreak and achieve teaching goals with online tools. However, he explained how it has also enabled teachers and students to try new ways to carry out teaching and research.

Year 4 Applied Psychology student Liang Haohai's research titled ‘Positive and Negative Peer Relationship among Preschool Children: Social Network Analysis’ won The Best Poster Award by Division and The Best Poster Award by Programme. The research results provided a new perspective for future studies on children's social networks and individual status.

Excerpt of Liang Haohai's research results

Chen Leyao, a Year 4 Statistics Programme student, won the Student Choice Award with her topic ‘Using News to Predict Stock Movements’ using deep learning methods. Content feature and sentiment feature extracted from the collected Weibo financial news were put into a Long-Short Term Memory model together with stock indexes, to predict stock movements. The experiment result implies that content and sentiment of news play an important role in predicting stock movements.

Part of Chen Leyao's poster

Guo Zhenyue and Lin Zhehan from Year 4 Computer Science and Technology studied various sentiment analysis technologies that are widely used in natural language processing as well as developing a chat room with sentiment analysis using the BERT model.

Interface of the chat room

Year 4 Environmental Science students Xiao Wanqi and Liang Lingzi studied the interspecific interaction between benthic invertebrates and halophila beccarii in the Tangjia area of Zhuhai. They evaluated the impact of human activities on the seagrass beds as they provide habitat for marine life and have important ecological functions and economic value.

The students found that the coverage of seagrass in Tangjia in 2019 decreased by 12.6% compared with 2017. Intensive human activities on the seagrass bed pose a threat to the growth of seagrass. This study revealed for the first time the ecological function of the seagrass community in Tangjia from the perspective of the interaction between the seagrass community and dominant species of invertebrates and clarifies the protection value of the seagrass bed.

Year 4 Food Science and Technology student Zhao Yingshu's study was ‘Konjac Glucomannan (KGM) Exerts Anti-diabetic Effects through Regulating Oxidative Stress and Inflammation’. In her presentation, she illustrated that hyperglycemia links with inflammation and oxidative stress. Her study found that anti-diabetic effects of KGM on rats with Type 2 Diabetes. The treatment with KGM has considerably reduced plasma glucose and insulin levels reduced the oxidative stress and Inflammation, positively regulated the pathways of Nrf2 and NF-kB, and helped protect the tissue structures.

The results and discussions of Zhao Yingshu's project

Year 4 Financial Mathematics student, Wang Zehua, designed a ‘Curiosity Based Recommendation System (CBRS)’. It differentiates from the traditional recommendation system which usually recommends similar content for the users.

The CBRS adopts the famous Wundt Curve theory in psychology, uses novelty to reverse quantify curiosity, and generates personalised recommendations based on user curiosity and relevance to users. As a result, more novel items that meet the user's preferences would be recommended.

Wang Zehua's research results

Year 4 Computer Science and Technology students, Yu Zhongyi, Wu Zhenghao and Zheng Hao, developed a model that can generate natural language summary with a preset number of words while preserving the main ideas of the original document. They noticed that all the existing models have to compromise between summary quality and length controllability. To address this problem, they proposed the novel-length controlling unit "Length Attention", and managed to achieve a good result.

Reporter: Covee Wang
Editors: Samuel Burgess, Deen He
(from MPRO)