Novel Use of Social Media Big Data and Artificial Intelligence for Community Resilience Assessment (CRA) in University Towns

Abstract

University towns face many challenges in the 21st century due to urbanization, increased student population, and higher educational institutions’ inability to house all their students on-campus. For university towns to be resilient and sustainable, the challenges facing them must be assessed and addressed. To carry out community resilience assessments, this study adopted a novel methodological framework to harness the power of artificial intelligence and social media big data (user-generated content on Twitter) to carry out remote studies in six university towns on six continents using Text Mining, Machine Learning, and Natural Language Processing. Cultural, social, physical, economic, and institutional and governance community challenges were identified and analyzed from the historical big data and validated using an online expert survey. This study gives a global overview of the challenges university towns experience due to studentification and shows that artificial intelligence can provide an easy, cheap, and more accurate way of conducting community resilience assessments in urban communities. The study also contributes to knowledge of research in the new normal by proving that longitudinal studies can be completed remotely.

A Composite Resilience Index (CRI) for Developing Resilience and Sustainability in University Towns

“…The proposed CRI is expected to contribute to the holistic measurement of community resilience in Akoka, it will help to minimize bias in assessing the level of resilience, help to develop resilience in Akoka against the negative impacts of studentification, provide a methodology for other university towns to develop their own CRI, and generally contribute to the resilience body of knowledge. This study also lays the foundation of future research combining AI and textual big data [40] and multicriteria decision-making tools to develop indices for community resilience and sustainability, beyond studentification.…”

“…Since this framework is recent and has a high accuracy level, there is no need to duplicate the effort here. Abdul-Rahman [39] used the framework to mine textual big data from six university towns, including Akoka, from 1 January 2010 to 31 December 2020 (10 years). This data, in its pre-processed form, are available online in a repository via Abdul-Rahman [39] and also Abdul-Rahman et al [40].…”

“…Abdul-Rahman [39] used the framework to mine textual big data from six university towns, including Akoka, from 1 January 2010 to 31 December 2020 (10 years). This data, in its pre-processed form, are available online in a repository via Abdul-Rahman [39] and also Abdul-Rahman et al [40]. These contain 935,822 user-generated contents (Tweets).…”

“…Some of these visitors include former residents in the town. The data mining, pre-processing, topic modeling and sentiments analysis procedures and algorithms are published in Abdul-Rahman, Adegoriola, McWilson, Soyinka and Adenle [40] so there is no need to repeat them here. The data extracted for this study are presented here in Table 1.…”

“…This helps during the periodic review of the implementation of the CRI in the town and allows fewer performing elements to be adjusted or upscaled [98]. The outcome indicators were developed from the mined opinions of the town's residents and visitors from Twitter [39] and prioritized by 23 experts (decision-makers). However, the process indicators were directly adopted from the ICBRR model [88,89].…”