Leveraging newspapers to understand urban issues

In the past, this blog has explored several aspects of Detroit, such as how well its covered with Volunteered Street View Imagery or how through the use of agent-based models one can explore issues with urban shrinkage. Keeping up with the theme of shrinkage and Detroit but at the same time utilizing our growing interest in natural language processing (especially topic modeling) we (Na (Richard) Jiang, Hamdi Kavak, Wenjing Wang and myself) have a new paper entitled "Leveraging newspapers to understand urban issues: A longitudinal analysis of urban shrinkage in Detroit" published in Environment and Planning B. 

In the paper, we take 6794 English news articles published by national and local press organizations (e.g., Forbes, The New York Times, Newsweek, The Detroit News) between 1975 to 2021 using the keywords “Detroit”, “shrink” and “decline.” These keywords were selected based on the characteristics of the study area (i.e., Detroit) and the phenomenon of urban shrinkage. With these data we then use BERTopic to detect and classify all collected news articles into certain topics. We chose BERTopic because it captures the semantic relationship among words converting sentences and words to embedding and automatically generates the topic unlike other NLP topic modeling techniques (e.g., LDA). Our topic modeling results identify several insights with respect to Detroit's shrinkage. For example, we can detect the side effects of the 2007-2009 economic recession on Detroit's automobile industry, local employment status, and the housing market. If sounds of interest and you want to find out more, below we provide the abstract, some figures from the paper including the methodology workflow and an example of the resulting topics over time. Finally, at the bottom of the page you can see the full reference and s link to the paper itself.

Abstract 

Today we are awash with data, especially when it comes to studying cities from a diverse data ecosystem ranging from demographic to remotely sensed imagery and social media. This has led to the growth of urban analytics providing new ways to conduct quantitative research within cities. One area that has seen significant growth is using natural language processing techniques on text data from social media to explore various issues relating to urban morphology. However, we would argue that social media only provides limited insights when dealing with longer-term urban phenomena, such as the growth and shrinkage of cities. This relates to the fact that social media is a relatively recent phenomenon compared to longer-term urban problems that take decades to emerge. Concerning longer-term coverage, newspapers, which are increasingly becoming digitized, provide the possibility to overcome the limitations of social media and provide insights over a timeframe that social media does not. To demonstrate the utility of newspapers for urban analytics and to study longer-term urban issues, we utilize an advanced topic modeling technique (i.e., BERTopic) on a large number of newspaper articles from 1975 to 2021 to explore urban shrinkage in Detroit. Our topic modeling results reveal insights related to how Detroit shrinks. For example, side effects of 2007 to 2009 economic recessions on Detroit’s automobile industry, local employment status, and the housing market. 

Key Words: Natural Language Processing, Topic Modeling, Newspapers, Urban Shrinkage, Urban Analytics.

 

 Vacancy status change from 1970 to 2010 for city of Detroit and surrounding area.

Topic modeling work flow.

Topics over time (a) urban, (b) population, (c) shrinkage, (d) economy, (e) job, (f) house.

Full Reference:

Jiang, N., Crooks, A.T., Kavak, H. and Wang, W. (2023), Leveraging Newspapers to Understand Urban Issues: A Longitudinal Analysis of Urban Shrinkage in Detroit, Environment and Planning B. Available at https://doi.org/10.1177/23998083231204695. (pdf)

Spatial Data Science Symposium

The other week Yingjie Hu and myself co-organized a session entitled "Spatial Data Science for Disaster Resilience" as part for the 4th Spatial Data Science Symposium (SDSS 2023). 

Session Abstract: 

Natural disasters, such as hurricanes, floods, tornados, wildfires, earthquakes, and blizzards, pose significant threats to people and society. The availability of various geospatial data sources (e.g., drone-collected images, mobile phone location data, social media data, and sensor network data) combined with the advancement of statistical and machine learning models provide great opportunities for understanding human-environment interactions during these catastrophic events. This session aims to bring together researchers interested in using spatial data science to answer questions and address issues in any aspect related to disaster management.

Talks in the session: 

• Lei Zou (keynote): 
• Achieving a Smart and Resilient Future with Spatial Data Science.
• Qunying Huang: 
• Wildfire Burnt Area Detection with Deep Learning and Sentinel2 Imagery.
• Manzhu Yu: 
• Deciphering Wildfire Dynamics: Spatiotemporal Attention-Based Sequence-to-Sequence Models Using ConvLSTM Networks.
• Md Zakaria Salim: 
• Socio-economic Disparities of Property Damage in Hurricane Ian.
• Qingqing Chen: 
• Community Resilience to Wildfire: A Network Analysis Approach by Utilizing Human Mobility Data.
• Kai Sun: 
• GALLOC: a GeoAnnotator for Labeling LOCation Descriptions from Disaster-related Text Messages.

If these talks sound of interest and as this was a online and distributed event, the main organizers of the Symposium have made all the talks available online.  The talks from our session can be seen below and all the other talks and seasons from the symposium at large can be found here.

 

Call for Abstracts: Geosimulations for Addressing Societal Challenges

As part of the The 10th Anniversary Symposium on Human Dynamics Research which will take place at the 2024 American Association of Geographers (AAG) Annual Meeting in Honolulu, Hawaii  between Tuesday, April 16 – Saturday, April 20, 2024 we are organizing a session(s) on Geosimulations for Addressing Societal Challenges. If the session description is of interest, please feel free to submit an abstract (details are below).

Session Description:

There is an urgent need for research that promotes sustainability in an era of societal challenges ranging from climate change, population growth, aging and wellbeing to that of pandemics. These need to be directly fed into policy. We, as a Geosimulation community, have the skills and knowledge to use the latest theory, models and evidence to make a positive and disruptive impact. These include agent-based modeling, microsimulation and increasingly, machine learning methods. However, there are several key questions that we need to address which we seek to cover in this session. For example, What do we need to be able to contribute to policy in a more direct and timely manner? What new or existing research approaches are needed? How can we make sure they are robust enough to be used in decision making? How can geosimulation be used to link across citizens, policy and practice and respond to these societal challenges? What are the cross-scale local trade-offs that will have to be negotiated as we re-configure and transform our urban and rural environments? How can spatial data (and analysis) be used to support the co-production of truly sustainable solutions, achieve social buy-in and social acceptance? And thereby co-produce solutions with citizens and policy makers.

We are particularly interested in presentations that will discuss issues relating to:

• Agent-based modeling and microsimulation techniques for responding to societal challenges; Agent-based models used for policy formation;
• Data driven modeling;
• Utilizing machine modeling for geosimulation;
• Creating really big models using exascale computation;
•  Model validation and assessment; 
• Participatory methods for agent-based modeling;
• Approaches to connect and share (open source) data and models;
• Revealing, quantifying, and reducing socio-economic inequalities with Geosimulation.

Next Steps:

If this sounds of interest, please e-mail the abstract and key words with your expression of intent to Richard Jiang (njiang8@buffalo.edu) by November 9th (one week before the AAG session deadline). Please make sure that your abstract conforms to the AAG guidelines in relation to title, word limit and key words and as specified at: https://aag.secure-platform.com/aag2024/page/abstracts/abstract-guidelines

An abstract should be no more than 250 words that describe the presentation’s purpose, methods, and conclusions. 

Timeline:

• 9th November, 2023: Abstract submission deadline. E-mail Richard Jiang by this date if you are interested in being in this session. Please submit an abstract and key words with your expression of intent.
• 14th November, 2023: Session finalization and author notification
• 15th November, 2023: Final abstract submission to AAG, via https://aag.secure-platform.com/aag2024/. All participants must register individually via this site. Upon registration you will be given a participant number (PIN). Send the PIN and a copy of your final abstract to Richard Jiang. Neither the organizers nor the AAG will edit the abstracts. 
• 16th November, 2023: AAG registration deadline. Sessions submitted to AAG for approval.
• 16th -20th April 2024: AAG in Honolulu.
  

Organizers

• Alison Heppenstall, University of Glasgow, Scotland.
  
• Na (Richard) Jiang, University at Buffalo, USA.
• Gary Polhill, The James Hutton Institute, Scotland.
• Andrew Crooks, University at Buffalo, USA.
• Raja Sengupta, McGill University, Canada.
  
• Suzana Dragicevic, Simon Fraser University, Canada. 
• Sarah Wise, University College London, England. 
• Jeon-Young Kang, Kyung Hee University, South Korea.
  

Agent-Based Modeling of Consumer Choice

At the upcoming International Conference on Geographic Information Science (GIScience 2023) Boyu Wang and myself have a new paper entitled "Agent-Based Modeling of Consumer Choice by Utilizing Crowdsourced Data and Deep Learning." In the paper we explore how through mining Yelp reviews can inform an agents choices of restaurants. The model itself was created in Mesa and uses Mesa-Geo and  more details about the model can be found at https://github.com/wang-boyu/yelp-abm.  If this sounds of interest, below you can see the abstract to the paper, some fugues including the graphical user interface of the model and a link to the paper.

Abstract: People’s opinions are one of the defining factors that turn spaces into meaningful places. Online platforms such as Yelp allow users to publish their reviews on businesses. To understand reviewers' opinion formation processes and the emergent patterns of published opinions, we utilize natural language processing (NLP) techniques especially that of aspect-based sentiment analysis methods (a deep learning approach) on a geographically explicit Yelp dataset to extract and categorize reviewers' opinion aspects on places within urban areas. Such data is then used as a basis to inform an agent-based model, where consumers' (i.e., agents') choices are based on their characteristics and preferences. The results show the emergent patterns of reviewers' opinions and the influence of these opinions on others. As such this work demonstrates how using deep learning techniques on geospatial data can help advance our understanding of place and cities more generally.

Keywords: aspect-category sentiment analysis, consumer choice, agent-based modeling, online restaurant reviews.

An overview of proposed agent-based model logic.

Average star rating vs. average sentiment by aspect category for 200 randomly selected restaurants in the City of St. Louis, MO.

The prototype agent-based model (a) with simulated (b) and actual visiting patterns (c).

Full reference:

Wang, B. and Crooks, A.T. (2023), Agent-Based Modeling of Consumer Choice by Utilizing Crowdsourced Data and Deep Learning, in Beecham, R., Long, J.A., Smith, D., Zhao, Q., and Wise, S (eds), Proceedings of the 12th International Conference on Geographic Information Science (GIScience 2023), Dagstuhl Publishing, Dagstuhl, Germany., pp. 81:1-81:6. (pdf)

ABM Online Courses

Often I get asked about how to learn about agent-based modeling (ABM). While we have a book on this with respect to GIS and ABM, the other day, Jiaqi Ge posted a question about free ABM online courses on the SIMSOC mailing list and I though it would be worth summarizing the responses here as the resources are quite useful.

Jiaqi shared some really good resources like the Santa Fe Institutes "Introduction to Agent-Based Modeling" and "Fundamentals of NetLogo" along with the University of Geneva's Coursera course "Simulation and modeling of natural processes". 

 

Others also responded to the question. For example, Wander Jager responded with online modules developed from the Action for Computational Thinking in Social Sciences (ACTiSS) team. Jen Badham responded with an extended tutorial about model design and creating models in Netlogo while Dino Carpentras responded with several general videos on YouTube on ABM which he has created. Hopefully readers will find these useful and also you might want to see our Github pages on GIS and ABM.