Name: Haoliang CHANG

Email: hlchang@ust.hk

GitHub: https://github.com/bright1993ff66

Date of Birth: 1993-02-17

Research Interests

• Social Media Data Analysis and Its Application in Transportation and Urban Planning
  • Traffic-related Social Media Data Detection, GIS analysis and Traffic Information Characterization
  • Social Media Community Detection and Profiling
• Natural Language Processing
  • Context Representation Learning (words, sentences, documents)
  • Application of Context Representation in Downstream Tasks such as Text Classification and Information Extraction

Education

• 2022.08 - Now
  Hong Kong University of Science and Technology, Guangzhou, China
  Postdoc in Urban Transportation Management and Public Policies
• 2018.09 - 2022.06
  City University of Hong Kong, Hong Kong, China
  PhD in Advanced Design and Systems Engineering
• 2015.09 - 2016.09
  The University of Edinburgh, Scotland, UK
  MSc. Statistics and Operational Research
  • Overall Classification of the Qualification: With Distinction
  • Average Points: 77/100
• 2011.09 - 2015.06
  Beifang University of Nationalities, Ningxia, China
  BSc. Statistics
  • Recipient of Chinese National Scholarship(2011-2012)
  • Average Points: 88/100

Publication

• Hao, T., Chang, H., Liang, S., Jones, P., Chan, P., Li, L., Huang, J. (2023). Heat and park attendance: Evidence from “small data” and “big data” in Hong Kong. Building and Environment, 234, 110123
• Chang, H., Li, L., Huang, J., Zhang, Q., & Chin, K. S. (2022). Tracking traffic congestion and accidents using social media data: A case study of Shanghai. Accident Analysis & Prevention, 169, 106618.
• Chang, H., Huang, J., Yao, W., Zhao, W., & Li, L. (2022). How do new transit stations affect people’s sentiment and activity? A case study based on social media data in Hong Kong. Transport Policy.
• Huang, J., Cui, Y., Chang, H., Obracht-Prondzyńska, H., Kamrowska-Zaluska, D., & Li, L. (2022). A city is not a tree: a multi-city study on street network and urban life. Landscape and Urban Planning, 226, 104469.

Research Service

• Reviewer, Accident Analysis and Prevention

Research Experience

• 2020.5 - Now
  Traffic Relevant Weibo Detection and Analysis in Shanghai
  • Use Weibo API to crawl the Weibos posted near Shanghai in 2012
  • Manually label the collected Weibos and train the deep learning models such as CNN-LSTM and Transformer based models to detect the traffic relevant Weibos
  • Based on the detected traffic relevant microblogs, conduct traffic relevant Weibo analysis, characterize the areas with high density of traffic events through text mining
• 2020.3 - Now
  Traffic Information Recommendation Using Graph Neural Networks
  • Use Twitter API to collect tweets posted in major US cities in 2018
  • Based on user’s historical interaction with traffic relevant messages in social media, build the graph between the social media users and places
  • Use Graph Attention Network to encode the user-place graph and make traffic information recommendation
• 2018.11 - 2020.5
  Evaluated the Influence of New Transit Stations on Nearby People in Hong Kong
  • Used Twitter API to collect tweets posted in Hong Kong from May 2016 to December 2018
  • Run ArcGIS to find the tweets posted in the walkable distance (500 meter) around the new transit stations
  • Applied sentiment analysis to estimate the sentiment of tweets posted near transit stations. Evaluated the influence of new transit stations on nearby people by comparing the sentiment and number of posted tweets before and after the introduction of transit stations

Project Experience

• 2018.10 - Now
  Crawl and Manage the Tweets Posted in 27 Cities Worldwide
  • Set up Amazon EC2 instances and used Twitter Streaming API to collect the tweets posted in major cities worldwide, including US cities(Atlanta, Boston, Chicago, etc.), European Cities(London, Madrid, etc.), and other major cities(Taipei, Bangkok, Tokyo, etc)
  • Monitor the tweet collection process on a daily basis
  • Saved the collected tweets to Amazon S3 and local server for the following research
• 2020.8-2020.9
  Find the Tweets Posted in Singapore Open Space
  • Crawled the land use data from OpenStreetMap in Singapore
  • Run Python scripts to get tweets with geo-information
  • Used the Spatial Join method in arcpy to find the tweets posted in Singapore open space such as amusement parks, playgrounds and green space

Work Experience

• 2016.10 - 2017.04 Research Executive
  Department of Automobile, Ipsos Beijing, China
  • Specialized in the research of whole automobile market and market segment, the evaluation of product and its positioning, the analysis of competitors, etc.
  • Used popular Python based modules such as nltk, gensim, and sklearn to do sentiment analysis of customers’ web comments on popular Chinese car brands
  • Used R packages (rvest, RCurl, etc.) to design web crawler from scratch and got useful raw data. Completed data visualization tasks using popular R packages(ggplot2, corrplot, etc.)
  • Participated in the on-site market research project, designed the questionnaire, discussed with car buyers about the automobile market and car models, and wrote the project report for clients from automobile industry

Skills

• Language: Chinese(Native Speaker), English, Spanish(Basic)
• Programming Languages: Python, R, SQL
• Framework: Tensorflow, PyTorch, NLTK, SpaCy, Gensim, DGL, networkx
• GIS Analysis: ArcGIS, python arcpy, geopandas
• Productivity Tools: LaTex, Markdown, Git, Vim, PyCharm, MySQL, Docker

Person

• Write articles about PhD application on Jianshu