HerWay: A Data-Driven System for Chicago Neighborhood Awareness
DOI:
https://doi.org/10.18409/jz73d329Keywords:
urban safety, sentiment analysis, named entity recognition, women's safety, multi-source integration, Chicago, equity, NLP, chatbot, Interactive MapAbstract
Urban safety is not experienced equally. Women, international students, and newcomers often lack the contextual knowledge needed to navigate public spaces confidently, creating an equity gap in information access that existing tools—crime maps and generic navigation apps—do not address. We present HerWay, a data-driven urban safety platform for Chicago integrating three complementary sources: Reddit community posts (1,218 posts, 2010–2026), Chicago Police Department crime records (200,000+ incidents, 2025), and City of Chicago 311 service requests. Using transformer-based sentiment analysis, gazetteer-based named entity recognition, and a weighted multi-source scoring framework, we produce neighborhood-level safety profiles for all 77 Chicago community areas. Key findings: female-perspective posts express fear at 28.5% vs. 24.2% overall, with night-time female fear reaching 33.3%—a 13.8 percentage-point gap that official crime data cannot capture. A novel source agreement signal surfaces neighborhoods where community perception and official data diverge. The platform is deployed as an interactive map and GPT-4o–powered chatbot at https://her-way-soremo.vercel.app.
References
R. Solymosi, B. Bowers, and K. Fujiyama, "Mapping fear of crime as a context-dependent everyday experience," Legal and Criminological Psychology, vol. 23, no. 2, pp. 116–135, 2018. https://doi.org/10.1111/lcrp.12108
H. Wang, D. Kifer, C. Graif, and Z. Li, "Crime rate inference with big data," in Proc. 22nd ACM SIGKDD, 2016, pp. 635–644. https://doi.org/10.1145/2939672.2939736
M. Hasan et al., "Detecting safety-relevant social media posts using BERT-based models," Information Processing & Management, vol. 59, no. 3, 2022. https://doi.org/10.1016/j.ipm.2022.102895
F. Barbieri, J. Camacho-Collados, L. Espinosa-Anke, and L. Neves, "TweetEval: Unified benchmark and comparative evaluation for tweet classification," in Findings of EMNLP 2020, pp. 1644–1650. https://doi.org/10.18653/v1/2020.findings-emnlp.148
Y. Liu et al., "RoBERTa: A robustly optimized BERT pretraining approach," arXiv:1907.11692, 2019. https://arxiv.org/abs/1907.11692
A. Vaswani et al., "Attention is all you need," in Adv. Neural Inf. Process. Syst., vol. 30, 2017. https://arxiv.org/abs/1706.03762
H. Koskela and R. Pain, "Revisiting fear and place: Women's fear of attack and the built environment," Geoforum, vol. 31, no. 2, pp. 269–280, 2000. https://doi.org/10.1016/S0016-7185(99)00033-0
K. Lum and W. Isaac, "To predict and serve?" Significance, vol. 13, no. 5, pp. 14–19, 2016. https://doi.org/10.1111/j.1740-9713.2016.00960.x
T. Gebru et al., "Using deep learning and Google Street View to estimate the demographic makeup of neighborhoods," Proc. National Academy of Sciences, vol. 114, no. 50, pp. 13108–13113, 2017. https://doi.org/10.1073/pnas.1700035114
City of Chicago, "Chicago crime data," Chicago Data Portal, 2024. https://data.cityofchicago.org/Public-Safety/Crimes-2001-to-Present/ijzp-q8t2
City of Chicago, "311 service requests," Chicago Data Portal, 2024. https://data.cityofchicago.org/Service-Requests/311-Service-Requests/v6vf-nfxy
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Sanjana Waghray, Gayathri Ananya Bhooplam Praveen , Sharanya Mishra
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
