Designing a natural language processing-driven communication system for urban planning: A case study

Abstract

Urban planning primarily depends on public input. However, citizen engagement is often limited by traditional communication channels. Historically, urban planning decisions have lacked mechanisms for incorporating feedback from the general public. Modern technology – particularly artificial intelligence (AI) – now offers the means to address this gap. Social media platforms, in particular, provide opportunities for broader and more dynamic interactions between citizens and planners. This study presents the design of an AI-based communication system for urban planning that gathers and analyzes citizen input in the Albanian language from social media platforms. Given the absence of existing natural language processing (NLP) tools for Albanian, the system uses a custom-built NLP pipeline, incorporating manual data preprocessing steps such as tokenization, lemmatization, and sentiment analysis. By integrating these preprocessing procedures with machine learning models for trend analysis and opinion classification, the system empowers urban planners to make data-driven decisions based on real-time public feedback. This work demonstrates the potential for additional automation and highlights the adaptability of AI techniques in addressing language-specific resource constraints.

Keywords

Machine learning / Communication system / Natural language processing / Social media data / Albanian language / Algorithms

Acknowledgments

None.

Funding

None.

Conflict of interest

The authors declare that they have no competing interests.

Author contributions

Conceptualization: All authors Formal analysis: All authors Investigation: All authors Methodology: All authors Writing – original draft: All authors Writing – review & editing: All authors

Ethics approval and consent to participate

Ethical approval was not obtained, as the study involved anonymous responses collected through Google Forms without any personal or identifiable data. The questionnaire clearly stated that participation was voluntary and that the data would be used solely for academic research. Given the nature of the study, we understood that formal ethical approval was not required. Informed consent was obtained from all participants before they submitted their responses.

Consent for publication

The introductory page of the questionnaire clearly stated that participation was voluntary and that all responses would remain anonymous. It also explained that the data would be used solely for academic research and potential publication purposes. By choosing to proceed with the questionnaire, participants confirmed their understanding of and consent to these terms.

Availability of data

The datasets used in this study are not publicly available, as they contain user-generated content collected via Google Forms distributed through social media. Due to privacy concerns and the absence of explicit participant consent for data sharing, the data cannot be made openly accessible.

References

[1] Al-Kodmany, K. (2000). Public participation: Technology and democracy. Journal of Architectural Education, 53(4): 220-228. https://doi.org/10.1162/104648800564439

[2] Amado, M. P., Santos, C. V., Moura, E. B., & Silva, V. G. (2010). Public participation in sustainable urban planning. International Journal of Human and Social Sciences, 5(2): 102-108. https://doi.org/10.5281/zenodo.1085046

[3] Aninditya, A., Hasibuan, M. A., & Sutoyo, E. (2019). Text mining approach using TF-IDF and Naive Bayes for classification of exam questions based on cognitive level of Bloom’s taxonomy. In: 2019 IEEE International Conference on Internet of Things and Intelligence System (IoTaIS), pp. 112-117. https://doi.org/10.1109/IoTaIS47347.2019.8980428

[4] Arora, R. (n.d.). 9 advantages and 10 disadvantages of Naive Bayes algorithm. In: OpenGenus IQ: Learn Algorithms, DL, System Design. Available from: https://iq.opengenus.org/advantages-and-disadvantages-of-naive-bayes-algorithm [Last accessed: 2024 Oct 23].

[5] Bakllamaja, A., & Collaku, N. (2013). Albania: The linkage between infrastructure development and sector strategic objectives is crucial for achieving results and sustainability in the sector. Mediterranean Journal of Social Sciences, 4(3): 659-666. https://doi.org/10.5901/mjss.2013.v4n3p659

[6] Berisha, E., Napierała, T., & Leśniewska-Napierała, K. (n.d.). Report on Participatory Planning Approaches and Practice at EU Level . Italy: Politecnico di Torino. https://doi.org/10.13140/RG.2.2.15901.22240

[7] Binhuraib, T. (2020). NLP with CNNs. Medium Publication. Available from: https://towardsdatascience.com/nlp-with-cnns-a6aa743bdc1e [Last accessed: 2024 Oct 24].

[8] Cacciari, I., & Ranfagni, A. (2024). Hands-on fundamentals of 1D convolutional neural networks-A tutorial for beginner users. Applied Sciences, 14(18): 8500. https://doi.org/10.3390/app14188500

[9] Çakır, H. S., & Levent, V. E. (2021). Data mining the city: User demands through social media. ICONARP International Journal of Architecture and Planning, 9(2): 799-812. https://doi.org/10.15320/ICONARP.2021.181

[10] Çerpja, A., & Çepani, A. (2022). Computational modeling of morphology in Albanian language: The case of verbs. Innovation in Language Learning 2022 - Conference Proceedings, p. 5761. Available from: https://conference.pixel-online.net/library_scheda.php?id_abs=5761 [Last accessed: 2025 Feb 07].

[11] Coglianese, C., & Dahl, R. (1990). Democracy and its critics. Michigan Law Review, 88(7): 1662-1670. https://doi.org/10.2307/1289333

[12] Co-PLAN. (2017). Annual Report 2017 . Available from: https://www.co-plan.org/wp-content/uploads/2021/01/co-plan-annual-report-2017.pdf [Last accessed: 2025 Feb 28].

[13] Cornwall, A. (2008). Unpacking “participation”: Models, meanings and practices. Community Development Journal, 43(3): 269-283. https://doi.org/10.1093/cdj/bsn010

[14] DataMiracle. (2024). Understanding the Classification Report in Machine Learning: A Comprehensive Guide . DataMiracle. Available from: https://datamiracle.com/2024/08/12/machine-learning-classification-reporit [Last accessed: 2025 Feb 27].

[15] DeepOpinion. (n.d.). Data Labelling Guideline for Aspect-based Sentiment Analysis . DeepOpinion. Available from: https://help.deepopinion.ai/faqs/data-labelling-guideline-for-aspect-based-sentiment-analysis [Last accessed: 2024 Oct 23].

[16] Dhrami, K., Ramaj, E., Muka, R., Sokoli, I., & Gjika, A. (2024). Deliverable 4.1 - Report on Implemented RRI “Anchor” Initiative in the Kune-Vain-Tale Lagoon Wetland Ecosystem . Co-PLAN. https://www.co-plan.org/wp-content/uploads/2025/02/D4.1_wbc-rri-net.pdf [Last accessed: 2025 Feb 20].

[17] Dong, L., Wang, Z., & Zhou, Y. (2023). Public participation and the effect of environmental governance in China: A systematic review and meta-analysis. Sustainability, 15(5): 4442. https://doi.org/10.3390/su15054442

[18] Dong, X., & Lian, Y. (2021). A review of social media-based public opinion analyses: Challenges and recommendations. Technology in Society, 67: 101724. https://doi.org/10.1016/j.techsoc.2021.101724

[19] Emekci, Ş. (2022). Gathering public opinion about an architectural project: A text mining approach. OPUS Toplum Araştırmaları Dergisi, 19(47): 506-515. https://doi.org/10.26466/opusjsr.1122785

[20] Fahmy Amin, M. (2023). Confusion matrix in three-class classification problems: A step-by-step tutorial. Journal of Engineering Research, 7(1): e296718. https://doi.org/10.21608/erjeng.2023.296718

[21] Fang, Z., Yang, T., & Jin, Y. (2020). DeepStreet: A Deep Learning Powered Urban Street Network Generation Module (arXiv:2010.04365) . arXiv [Preprint]. https://doi.org/10.48550/arXiv.2010.04365

[22] Fillim. (n.d.). Available from: https://shkendija.org [Last accessed: 2025 Jan 10].

[23] GeeksforGeeks. (2020). What is Sentiment Analysis? GeeksforGeeks. Available from: https://www.geeksforgeeks.org/what-is-sentiment-analysis [Last accessed: 2024 Nov 20].

[24] GeeksforGeeks. (2024). Keras vs TensorFlow vs PyTorch . GeeksforGeeks. Available from: https://www.geeksforgeeks.org/keras-vs-tensorflow-vs-pytorch [Last accessed 2024 Feb 28].

[25] GeeksforGeeks. (n.d.). Text Classification Using CNN . GeeksforGeeks. Available from: https://www.geeksforgeeks.org/text-classification-using-cnn [Last accessed 2025 Feb 27].

[26] Ghahramani, M., Galle, N. J., Ratti, C., & Pilla, F. (2021). Tales of a city: Sentiment analysis of urban green space in Dublin. Cities, 119: 103395. https://doi.org/10.1016/j.cities.2021.103395

[27] Glue, R. (2023). TfIdf + Naive Bayes Beat Neural Nets . Available from: https://medium.com/@rustemgal/tfidf-naive-bayes-beat-neural-nets-a58f9dff3741 [Last accessed: 2024 Oct 23].

[28] Guo, J., & Bai, J. (2019). The role of public participation in environmental governance: Empirical evidence from China. Sustainability, 11(17): 4696. https://doi.org/10.3390/su11174696

[29] Hakiman, K., & Sheely, R. (2023). Unlocking the potential of participatory planning: How flexible and adaptive governance interventions can work in practice. Studies in Comparative International Development, 58(1): 1-38. https://doi.org/10.1007/s12116-023-09415-x

[30] Healey, P. (1992). Planning through debate: The communicative turn in planning theory. Town Planning Review, 62(2): 143-162. https://doi.org/10.3828/tpr.63.2.422x602303814821

[31] Heng, J., Liu, J., & Lim, K. H. (2020). Urban Crowdsensing Using Social Media: An Empirical Study On Transformer and Recurrent Neural Networks (arXiv:2012.03057) . arXiv [Preprint]. https://doi.org/10.48550/arXiv.2012.03057

[32] Jian, L. (2024). Infrastructure investment and urban development in China. Journal of Poverty, Investment and Development, 9: 37-49. https://doi.org/10.47604/jpid.2586

[33] Jurafsky, D., & Martin, J. H. (n.d.). Naive Bayes, text classification, and sentiment. Speech and Language Processing, 3(4): 60-94. Available from: https://web.stanford.edu/~jurafsky/slp3/old_feb24/ed3book_jan24.pdf [Last accessed: 2025 Feb 20].

[34] Karic, S. (2024). Citizen participation and acceptance in the context of urban events: An investigation on regional garden shows in Germany. Planning Practice and Research, 39(3): 479-497. https://doi.org/10.1080/02697459.2024.2315392

[35] Khurana, D., Koli, A., Khatter, K., & Singh, S. (2023). Natural language processing: State of the art, current trends and challenges. Multimedia Tools and Applications, 82(3): 3713-3744. https://doi.org/10.1007/s11042-022-13428-4

[36] Kiranyaz, S., Avci, O., Abdeljaber, O., Ince, T., Gabbouj, M., & Inman, D. J. (2021). 1D convolutional neural networks and applications: A survey. Mechanical Systems and Signal Processing, 151: 107398. https://doi.org/10.1016/j.ymssp.2020.107398

[37] Kote, N., Biba, M., Kanerva, J., Rönnqvist, S., & Ginter, F. (2019). Morphological Tagging and Lemmatization of Albanian: A Manually Annotated Corpus and Neural Models . arXiv [Preprint]. https://doi.org/10.48550/arXiv.1912.00991

[38] Kowalski, R., Esteve, M., & Jankin Mikhaylov, S. (2020). Improving public services by mining citizen feedback: An application of natural language processing. Public Administration, 98(4): 1011-1026. https://doi.org/10.1111/padm.12656

[39] Kromydas, B. (2023). Convolutional Neural Network (CNN): A Complete Guide . LearnOpenCV. Available from: https://learnopencv.com/understanding-convolutional-neural-networks-cnn [Last accessed: 2024 Oct 23].

[40] Kurani, A., & Muho, A. (2014). A morphological comparative study between Albanian and English language. European Scientific Journal, 14. https://doi.org/10.19044/esj.2011.v14n0p%p

[41] Lijphart, A. (1997). Unequal participation: Democracy’s unresolved dilemma. The American Political Science Review, 91(1): 1-14. https://doi.org/10.2307/2952255

[42] Lopez, E., Abascal-Mena, R., & Zepeda-Hernández, S. (2017). Social media participation in urban planning: A new way to interact and take decisions. In: ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLII-4/W3, pp. 59-64. https://doi.org/10.5194/isprs-archives-XLII-4-W3-59-2017

[43] Mondal, S., & Goswami, S. S. (2024). Machine learning techniques for quality assurance in additive manufacturing processes. International Journal of AI for Materials and Design, 1(2): 21-40. https://doi.org/10.36922/ijamd.3455

[44] Morais, M. (2022). Citizen Participation in Urban Policy: Lessons Based on Berlin and São Paulo Experiences . WZB Discussion Paper, SP V 2022-101. Available from: https://ideas.repec.org/p/zbw/wzbdsc/spv2022101.html [Last accessed: 2024 Nov 20].

[45] Murray, M., Greer, J., Houston, D., McKay, S., & Murtagh, B. (2009). Bridging top down and bottom up: Modelling community preferences for a dispersed rural settlement pattern. European Planning Studies, 17(3): 441-462. https://doi.org/10.1080/09654310802618101

[46] Najafabadi, M. K. (2024). Sentiment analysis incorporating convolutional neural network into hidden Markov model. Computational Intelligence, 40(2): e12633. https://doi.org/10.1111/coin.12633

[47] Navarro, N., & Renaud, L. (2022). Forms of public participation in mobile apps for visitors: From user representation profile to citizen representation profile. Hybrid - Revue Des Arts et Médiations Humaines , 8: 8. https://doi.org/10.4000/hybrid.2010

[48] Newatia, R. (n.d.). How to Implement CNN for NLP Tasks Like Sentence Classification . Medium Publication. Available from: https://medium.com/saarthi-ai/sentence-classification-using-convolutional-neural-networks-ddad72c7048c [Last accessed: 2025 Feb 27].

[49] Nilsson, A., Vanhuyse, F., Arra, V., Requena, A., & Agerström, M. (2021). Citizen Participation for Sustainable Cities . Stockholm Environment Institute. Available from: https://www.sei.org/wp-content/uploads/2021/05/factsheet-citizen-participation-for-sustainable-cities.pdf [Last accessed: 2025 Feb 18].

[50] Ninja, N. (2023). TF-IDF: Weighing Importance in Text . Let’s Data Science. Available from: https://letsdatascience.com/tf-idf [Last accessed: 2024 Oct 23].

[51] Omrany, H., & Al-Obaidi, K. M. (2024). Application of digital twin technology for Urban Heat Island mitigation: Review and conceptual framework. Smart and Sustainable Built Environment, 13(1): 85-116. https://doi.org/10.1108/SASBE-05-2024-0189

[52] Ouyang, Y., & Bai, X. (2025). Social media for public participation in urban planning in China based on place attachment-A case of the Guangzhou banyan tree incident. Frontiers in Built Environment, 10: 1523576. https://doi.org/10.3389/fbuil.2024.1523576

[53] Pogrebinschi, T. (2021). Codebook for the LATINNO Dataset: Technical Report . WZB Discussion Paper, SPV2021-101. Available from: https://bibliothek.wzb.eu/pdf/2021/v21-101.pdf

[54] Qian, K., Mao, L., Liang, X., Ding, Y., Gao, J., Wei, X., et al. (2023). AI Agent as Urban Planner: Steering Stakeholder Dynamics in Urban Planning Via Consensus-based Multi-agent Reinforcement Learning . arXiv:2310.16772. [ arXiv Preprint]. https://doi.org/10.48550/arXiv.2310.16772

[55] Raza, A. (2024). Convolutional Neural Networks (CNNs) for Text Classification . Medium Publications. Available from: https://medium.com/@aliraza.abro.prog/convolutional-neural-networks-cnns-for-text-classification-bd473c7285a4 [Last accessed: 2025 Feb 27].

[56] Roul, R. K., Sahoo, J. K., & Arora, K. (2017). Modified TF-IDF term weighting strategies for text categorization. In: 2017 14^th IEEE India Council International Conference (INDICON) , pp. 1-6. https://doi.org/10.1109/INDICON.2017.8487593

[57] Scikit-learn. (n.d.). 1.9. Naive Bayes. Scikit-learn. Available from: https://scikit-learn.org/stable/modules/naive_bayes.html [Last accessed: 2024 Oct 23].

[58] Shao, P., & Wang, Y. (2016). How does social media change Chinese political culture? The formation of fragmentized public sphere. Telematics and Informatics, 34: 92-103. https://doi.org/10.1016/j.tele.2016.05.018

[59] Shehu, D., & Lezzerini, L. (2024). The challenges of applying big data in the urban planning practices for the developing countries: Case study in Albania. In: Proceedings of the International Conference on Architecture and Urban Planning, p. 218. https://doi.org/10.37199/c41000117

[60] Shehu, D., & Luarasi, T. (2024). A review on citizen motivation analyses for participating on urban planning process. In: Proceedings of the International Conference on Architecture and Urban Planning, p. 55. https://doi.org/10.37199/c41000706

[61] Shelf. (2024). Challenges and Considerations in Natural Language Processing . Shelf. Available from: https://shelf.io/blog/challenges-and-considerations-in-nlp [Last accessed: 2024 Oct 24].

[62] Swimm Team. (n.d.). 5 Types of Word Embeddings and Example NLP Applications . Swimm. Available from: https://swimm.io/learn/large-language-models/5-types-of-word-embeddings-and-example-nlp-applications [Last accessed: 2024 Oct 23].

[63] Tong, S., & Li, S. (2024). Explainable artificial intelligence for urban planning: Challenges, solutions, and future trends from a new perspective. International Journal of Advanced Computer Science and Applications, 15(7): 77-84. https://doi.org/10.14569/IJACSA.2024.0150777

[64] United Nations Human Settlements Programme . (2011). Cities and Climate Change: Policy Directions: Global Report on Human Settlements 2011 . Oxford, UK: Earthscan. Available from: https://unhabitat.org/sites/default/files/download-manager-files/Cities%20and%20Climate%20Change%20Global%20Report%20on%20Human%20Settlements%202011%20-%20ABRIDGED.pdf [Last accessed: 2025 Feb 05].

[65] United Nations. (1992). Agenda 21: Programme of Action for Sustainable Development . United Nations . Available from: https://sustainabledevelopment.un.org/content/documents/agenda21.pdf [Last accessed: 2025 Feb 05].

[66] Van Otten, N. (2023). CNN in NLP: Top 5 Types and How-to Tutorial for Text Classification with TensorFlow. Spot Intelligence. Available from: https://spotintelligence.com/2023/01/10/cnn-in-nlp [Last accessed: 2025 Feb 27].

[67] Wei, W., Wang, L., Li, X., Zhang, B., & Scherer, R. (2021). Design and implementation of public opinion monitoring system based on cloud platform. Journal of Internet Technology, 22(3): 569-581. https://doi.org/10.3966/160792642021052203007

[68] Yang, R. (n.d.). Convolutional neural networks and their applications in NLP. In: Modern Approaches in Natural Language Processing . Available from: https://slds-lmu.github.io/seminar_nlp_ss20/convolutional-neural-networks-and-their-applications-in-nlp.html [Last accessed: 2024 Oct 23].

[69] Zhang, W. (2023). Hybrid Deliberation: Citizen Dialogues in a Post-pandemic Era . arXiv:2307.11412. [ arXiv Preprint]. https://doi.org/10.48550/arXiv.2307.11412