Leveraging heterogeneous data and machine learning for mapping connectivity and infrastructure provision in Amazon

Poverty is a complex and multidimensional issue that is difficult to measure accurately. While multiple studies have employed traditional econometric methods to analyze poverty, they often overlook the critical roles of electricity, Internet, and cell phone access – factors our study incorporates alongside machine learning to provide deeper and more accurate insights. This study examines the associations between Internet, cell phone, and electricity access and poverty in the poorest region of Luzon, Philippines, aiming to foster connectivity among households to support poverty alleviation. Using probit regression and estimation analyses, we found significant socioeconomic disparities, with many households living below the poverty line. While most households have electricity and cell phones, many lack Internet access. This indicates challenges in infrastructure and digital connectivity that affect living standards and economic opportunities. The analysis reveals important causal relationships between Internet access, cell phone ownership, electricity availability, household size, and the likelihood of being in poverty. Interestingly, the lack of these essential services is linked to higher poverty rates. These results highlight the need for targeted interventions to tackle the root causes of poverty, particularly in bridging the digital divide and improving access to essential services. Machine learning algorithms were employed to effectively predict poverty outcomes based on the results of econometric modeling, where variables with significant coefficients served as a priori inputs. The findings indicate that Extreme Gradient Boosting achieved the lowest mean square error and the highest R2 value among all regression models. Meanwhile, the random forest classifier demonstrated the best overall performance with the highest classification accuracy. The outlined policies support energy affordability, cell phone access, and Internet connectivity through financial aid, solar programs, device provisions, and broadband expansion. Addressing infrastructure gaps and technology access is key to sustainable economic growth, guiding policy makers toward equitable, resilient solutions.

Background: The COVID-19 pandemic has led to the rapid expansion of telehealth use. Telehealth has the potential to improve access for underserved populations who live far distances from well-equipped medical centers. This healthcare modality will likely be particularly important for patients with cancer. We hypothesized that there are disparities in the resources available to utilize telehealth and sought to study patient access and knowledge of associated technologies. Methods: We conducted a single-center cross-sectional survey study of patients at an NCCN-designated comprehensive cancer center over a two-month period. Demographics, education, internet, and cell phone access were assessed. Participant technological knowledge was determined with a validated 10-question quiz of terms regarding computers and the internet. Income was extrapolated using IRS data for individual zip codes. Analysis was performed using either an ordered logistic regression or mixed model ordered logistic regression (Stata SE 16.1). Results: There were 344 survey responses (rate 64.3%). Mean age of the respondents was 57.5 years, 70.0% were women, 67.3% were Caucasian, and 25.4% were Black. Many patients (35.0%) did not attend college, and 5.0% had not finished high school. The median estimated income was ≥45,820. Ninety-six patients (30.0%) did not have adequate internet access for telehealth use with 6.5% not having any internet access at all. Participants scored an average of 57.4% correct on the technology knowledge quiz. In unadjusted analysis, technological knowledge was predicted by age (Fig A), race (Fig B), income (Fig C), education level (Fig D), and cell phone type (Fig E). The type of home internet access and was not predictive of quiz score. Age, race, education level, and cell phone type remained significantly predictive of technological knowledge in multivariable analysis (Fig F). Home internet access (p =0.416) and median income (p =0.109) were not significantly associated with quiz score. Conclusions: In our sample, nearly one-third of cancer patients did not have adequate internet access for telehealth. There were significant disparities in technological knowledge among cancer patients which were associated with demographic characteristics. This study provides information for identifying patients that may have barriers to successful utilization of telehealth.

PurposeThe purpose of the study was to examine associations of gender role beliefs with marital safety as well as social and digital connectivity among married adolescent girls in India. MethodsWe analyzed cross-sectional survey data from married adolescent girls from rural Bihar and Uttar Pradesh, India in 2015–2016 (N = 4,893). Gender role belief items assessed participants' beliefs regarding appropriateness of female marital choice and economic decision-making, male childcare responsibility, and marital violence. The outcomes were time with friends (social connection), freedom of movement, mobile phone ownership and internet access (digital connection), and safety from marital violence (marital safety). Adjusted regression models examined associations between gender role beliefs and outcomes, caste/religion beliefs related to segregation, and demographics. ResultsBeliefs supportive of female marital choice (adjusted odds ratio [AOR] 1.38, 95% confidence interval [CI] 1.00–1.88, p = .048) and female economic decision-making (AOR 1.43, 95% CI 1.03–1.99, p = .03) were associated with social connection. Beliefs supportive of female marital choice (AOR 1.88, 95% CI 1.31–2.71, p = .001), female economic decision-making (AOR 1.67, 95% CI 1.03–2.72, p = .04), and male childcare responsibilities (AOR 1.42, 95% CI 1.05–1.94, p = .03) were associated with freedom of movement. Belief supporting female marital choice was associated with mobile phone ownership (AOR 1.23, 95% CI 1.01–1.50, p = .04), and belief supporting male childcare responsibility was associated with internet access (AOR 1.76, 95% CI 1.11–2.77, p = .02). Beliefs supportive of female marital choice (AOR .73, 95% CI .59–.89, p = .003), male childcare responsibility (AOR 1.26, 95% CI 1.03–1.54, p = .02), and unacceptability of marital violence (AOR 1.72, 95% CI 1.38–2.15, p < .001) were associated with safety from marital violence. ConclusionsProgressive gender role beliefs regarding female choice, control, and safety in marriage are associated with greater connectivity and safety for married adolescent girls in India. Future studies that can longitudinally assess these associations are warranted.

In this work we discussed core issues about maximise the potential of digital technologies to improve lives, addressed key topics, including new developments in the field of sustainable economy. Measurement is crucial for achieving universal and meaningful connectivity (UMC). Accurate and comprehensive data help us understand priorities, design effective interventions and monitor progress. Each country has its own characteristics, faces its own challenges, and must follow its own path to UMC. Low-income economies show notable advancements, driven by gains in Internet access, mobile broadband penetration, and affordability. To assist stakeholders in their research work, the ICT Development Index (IDI) was launched to give a measure of countries’ digital connectivity. It was revived in 2023, with a new methodology that has UMC at its core. The country reviews are an important tool for regulators and policymakers as they work to create an environment that promotes investment, competition, digital innovation, protects consumers, and ensures that the benefits of digital transformation are widely shared. They highlight diverse experiences and different regulatory patterns while exploring good practices, challenges and lessons learnt by regulators in navigating digital transformation. The country reviews also help develop a better understanding of the role and impact of collaboration and collaborative governance, and the use of new tools for regulating ICT markets. The review is based on ITU publications: “Egypt’s digital transformation and collaborative regulation” and “The ICT Development Index”.

Joining Zone 9

This study examines the impact of digital skills and internet access on the participation of citizens in the Republic of Kosovo in the digital economy. Data were collected using structured questionnaires from 326 respondents across various regions of Kosovo. Given the data structure, the study employed logistic regression analysis to investigate the relationship between digital skills, internet access and engagement in digital economic activities. The results indicate a significant positive relationship between internet access and digital skills with engagement in the digital economy. Specifically, individuals with higher digital skills and better internet access demonstrate higher levels of participation in digital economic activities. Furthermore, the study identifies demographic factors such as age and profession that significantly influence digital engagement, highlighting the need for targeted policy interventions to close the digital gap. These findings suggest that improving digital infrastructure and developing skills, especially to elders and rural living are crucial for enhancing engagement in digital economic activities. These insights provide a valuable overview of the current situation and reveal that even in emerging countries, access to the internet might be high. Moreover, can serve as guidance for policymakers aiming to foster a robust digital economy in Kosovo.

The rapid expansion of digital financial services has significantly transformed household financial behavior, particularly in developing economies. However, the effectiveness of these tools in improving household financial efficiency depends not only on their usage but also on users' digital literacy and access to supporting infrastructure. This study aims to examine the influence of digital payment system usage, digital financial literacy, and internet access on household financial efficiency in the context of Indonesia's digital economy. A quantitative survey was conducted among 120 households in Banyumas Regency, Central Java, using a structured questionnaire. Data were analyzed using multiple linear regression after validating assumptions of normality, multicollinearity, and heteroscedasticity. The variables measured included frequency of digital payments, financial literacy levels, internet accessibility, and self-reported financial efficiency. The results indicate that all three independent variables significantly and positively affect household financial efficiency, with digital financial literacy exerting the strongest influence. Digital payment usage and internet access also contributed significantly. These findings suggest that financial efficiency is not merely a function of access or usage but is strengthened when supported by adequate literacy and infrastructure. The study highlights the interdependent role of behavioral and structural factors in enabling effective household financial management. This research contributes to the growing literature on digital financial behavior by focusing on the household level in a semi-urban setting. The findings offer practical implications for policymakers, fintech developers, and financial educators aiming to promote inclusive and efficient digital financial ecosystems.

Loneliness partially mediates the relationship between Internet access and cognitive function in late adulthood: A cross-lagged panel analysis

Telemedicine (TM) has augmented healthcare by enabling remote consultations, diagnosis, treatment, and monitoring of patients, thereby improving healthcare access and patient outcomes. However, successful adoption of TM depends on user acceptance, which is influenced by technical, socioeconomic, and health-related factors. Leveraging machine learning (ML) to accurately predict these adoption factors can greatly contribute to the effective utilization of TM in healthcare. The objective of the study was to compare 12 ML algorithms for predicting willingness to use TM (TM try) among patients with rheumatic and musculoskeletal diseases (RMDs) and identify key contributing features. We conducted a secondary analysis of RMD patient data from a German nationwide cross-sectional survey. Twelve ML algorithms, including logistic regression, random forest, extreme gradient boosting (XGBoost), and neural network (deep learning) were tested on a subset of the dataset, with the inclusion of only RMD patients who answered “yes” or “no” to TM try. Nested cross-validation was used for each model. The best-performing model was selected based on area under the receiver operator characteristic (AUROC). For the best-performing model, a multinomial/multiclass ML approach was undertaken with the consideration of the three following classes: “yes”, “no”, “do not know/not answered”. Both one-vs-one and one-vs-rest strategies were considered. The feature importance was investigated using Shapley additive explanation (SHAP). A total of 438 RMD patients were included, with 26.5% of them willing to try TM, 40.6% not willing, and 32.9% undecided (missing answer or “do not know answer”). This dataset was used to train and test ML models. The mean accuracy of the 12 ML models ranged from 0.69 to 0.83, while the mean AUROC ranged from 0.79 to 0.90. The XGBoost model produced better results compared with the other models, with a sensitivity of 70%, specificity of 91% and positive predictive value of 84%. The most important predictors of TM try were the possibility that TM services were offered by a rheumatologist, prior TM knowledge, age, self-reported health status, Internet access at home and type of RMD diseases. For instance, for the yes vs. no classification, not wishing that TM services were offered by a rheumatologist, self-reporting a bad health status and being aged 60–69 years directed the model toward not wanting to try TM. By contrast, having Internet access at home and wishing that TM services were offered by a rheumatologist directed toward TM try. Our findings have significant implications for primary care, in particular for healthcare professionals aiming to implement TM effectively in their clinical routine. By understanding the key factors influencing patients' acceptance of TM, such as their expressed desire for TM services provided by a rheumatologist, self-reported health status, availability of home Internet access, and age, healthcare professionals can tailor their strategies to maximize the adoption and utilization of TM, ultimately improving healthcare outcomes for RMD patients. Our findings are of high interest for both clinical and medical teaching practice to fit changing health needs caused by the growing number of complex and chronically ill patients.

Addressing the global challenge of sustainable waste management, this research investigates the influence of social dynamics and digital connectivity on rural residents’ willingness to adopt waste classification practices, essential for sustainable environmental management. Through a comprehensive analysis of 5413 rural participants surveyed in the China Labor-force Dynamic Survey (CLDS), this study employs a novel mixed-methods approach. It integrates quantitative analysis with the Manski social interaction framework and a Recursive Bivariate Probit model to explore the intricate interplay between community interactions, internet access, and environmental behaviors. Our methodology stands out for its unique combination of social theory and econometric modeling to address a pressing environmental issue. Results highlight a significant effect of mobile internet use and social interactions within communities on enhancing willingness towards waste classification. Notably, digital connectivity emerges as a key facilitator of environmental engagement, mediating social influences, and fostering a collective approach to waste management. Considering these insights, we propose targeted policy interventions that blend digital strategies with traditional community engagement efforts. Recommendations include crafting digital literacy programs and leveraging social media to bolster community-centric environmental governance. By harnessing the synergistic potential of digital tools and social dynamics, these strategies aim to elevate the effectiveness of waste classification initiatives in rural China, offering a scalable model for environmental sustainability.

Rural areas continue to face digital inequality compared to urban areas. Urban areas have access to a myriad of next generation advanced information communications technology (ICT) whereas rural areas experience disparity of service type, price and reliability. Initially the urban-rural digital divide was one of quantity of subscribers or demand driven digital inclusion , it has now matured to become an issue of quality and capacity of connectivity. However, Silva, Badasyan & Busby found that the more rural a census tract is, the lower the broadband adoption rate. They found, in further testament to the need for increased capacity, that “broadband availability has the strongest impact on the adoption rate in non-metropolitan areas. If the availability were to increase to a 100%....it would increase the adoption rate by 6.12%”. Public policy and regulation has a direct impact upon availability or lack thereof for ICT services in rural areas. Hollifield, Donnermeyer, Wolford & Agunga found that when public policy, in the form of universal service funding in rural high cost areas, failed to support the early implementation of ICT services, communities began investing in self-development projects with limited success. This effort however, doesn’t address the circumstances of the rural household miles from an organized rural community. With the regulatory focus on increasing competition among ICT service providers since the passage of the Telecommunications Act of 1996, consumers in urban areas have benefited. In contrast, underserved or unserved rural consumers of ICT often pay a rural penalty in the form of a combination of one or more complicating factors, including; higher prices and lower bandwidth , and less reliability or no service at all. Whitacre & Mills reports several early studies pointed to the need for demand-oriented programs such as computer training, and demonstrations on internet usage, as more important than the development of access infrastructure to promote ICT usage in rural areas. However, as general societal demand for internet access grows from the diffusion of knowledge, infrastructure becomes the limiting factor in the development of rural areas. Lower population density areas are minimally (or un)profitable markets for ICT service providers and some areas may never experience for-profit investment in broadband provisioning due to the lack of a potential subscriber base. During the years of the traditional land-line telephone, high cost rural areas were supported by the federal universal service fund through subsidies paid by all users, rural and urban. This fund, reformulated as the Connect America Fund, was intended to take over this role in the new ICT economy, but has fell short of fulfilling its intended purpose. Additionally concerns exist with the measurement and reporting of the diffusion of broadband infrastructure. All facilities based service providers providing internet connection speeds exceeding 200 kbps must report bandwidth speeds to the Federal Communication Commission through form 477. Reporting of available bandwidth speeds initially occurred by zip code, and then by census tract, followed by a further adjustment to the level of census block, making longitudinal comparisons over multiple years difficult. Since 2014 broadband adoption rates are reported on the basis of a scale from 0 to 5 with 5 meaning over 800 broadband connections per 1,000 households. This is a problematic measurement methodology for low population density areas. Grubesic reports as of the first iteration of the national broadband map, the availability of broadband in the United States is overestimated. Both public policy and available technologies directly affect regional development. Salemink, Strijker & Bosworth has concluded that economic differences between “well-connected” and “poorly connected” areas will continue to grow. Mahasuweerachai, Whitacre, & Shideler concluded that rural counties with both digital subscriber line (DSL) and cable broadband technologies attracted a net positive number of in-migrations as compared to counties without broadband or with only one type of broadband. Salemink, Strijker & Bosworth further concludes that while not instrumental for economic growth in rural areas, digital connectivity is essential to support existing industries. Gallardo and Schmmahorn found that as the number of broadband providers increased, so did non-innovative entrepreneurs and as non-innovative entrepreneurs increased, income inequality decreased. A need exists for the accurate measurement and reporting of consumer available bandwidth to better understand this element of the urban-rural digital divide. The primary question for this research study is, can the rural-urban digital divide be accurately measured? To address this question and to better understand available ICT in a given region, two pilot studies have been performed in households to measure consumer available bandwidth and to ascertain multiple elements of consumer perceptions of their internet access. These pilot projects focused on two themes of digital connectivity issues as outlined by Salemink, Strijker & Bosworth including; policy and regulation, and technologies in rural areas.

Background: Remote and digitally disconnected communities experience significant health disparities due to limited access to healthcare, poor socioeconomic conditions, and lack of digital resources. These challenges contribute to higher rates of preventable diseases and delays in medical intervention. Objective: This study aimed to identify the risk factors associated with poor health outcomes in remote communities of Pakistan and assess the role of healthcare accessibility, digital connectivity, and socioeconomic constraints in worsening these disparities. Methods: A cross-sectional study was conducted from April 2024 to October 2024 in five rural areas of Khyber and Sindh, Pakistan, including Chitral, Upper Dir, Ghotki, Tharparkar, and Umerkot. A sample of 420 participants was selected using stratified random sampling. Data were collected through structured interviews, anthropometric measurements, and standardized health assessment tools. Statistical analysis was performed using SPSS version 26, employing chi-square tests and logistic regression to identify significant predictors of poor health outcomes. Ethical approval was obtained, and informed consent was secured from all participants. Results: The prevalence of hypertension, diabetes, malnutrition, anemia, and respiratory diseases was 35.7%, 28.6%, 42.9%, 33.3%, and 31.0%, respectively. Limited healthcare access was a major concern, with 45.2% residing over 10 km from a health facility, 59.5% lacking digital health resources, and 40.5% unable to afford medications. Digital connectivity remained low, with only 33.3% having internet access and 21.4% utilizing digital health services. Conclusion: Health disparities in remote communities are driven by financial constraints, inadequate healthcare facilities, and poor digital access. Strengthening healthcare infrastructure and digital inclusion strategies can mitigate these inequalities and improve health outcomes.

This research adopts a repertoire approach to examine the concept of a health information repertoire defined as a set of sources through which people get health information. Drawing on a random sample survey in Austin, TX, it borrows the concepts of cultural omnivores and univores to investigate how health information repertoire are related to social capital and digital inequalities. Results demonstrate that both the size and the composition of health information repertoires vary by social and digital connectivity. People with greater social capital have a larger repertoire and are less likely to be univores dependent on the Internet or interpersonal contacts. People with Internet access have a larger repertoire and are less likely to be univores dependent on television. More skilled Internet users are less likely to be univores dependent on interpersonal contacts, whereas frequent Internet users are more likely to be omnivores with a four‐channel repertoire including the Internet, interpersonal contacts, television, and newspaper. The positive relationship between social capital and repertoire size is stronger among less‐skilled Internet users. There are significant variations in health information repertoires in terms of media access and sociodemographic characteristics. Scholarly and practical implications are discussed.

ObjectiveThe introduction of the “Broadband China” strategy significantly boosted internet access for older adults in China. This study examines the association between improved internet usage on the psychological well-being of older adults in China and explores the mediating role of intergenerational interactions, focusing on economic and emotional exchanges between older adults and their children.MethodA difference-in-differences analysis was employed to evaluate the effectiveness of the “Broadband China” strategy in enhancing internet access. Additionally, fixed-effect regressions were used to examine the relationship between internet use and psychological well-being for older adults (aged 60 and above, male and female, all of Chinese ethnicity).ResultsInternet use is positively associated with psychological well-being. Mobile use and concurrent use of mobile and computers devices show particularly strong benefits. An inverted U-shaped relationship is observed between usage duration and well-being. Moderate use enhances well-being while excessive use diminishes the effect and may even lead to negative outcomes. Internet-based learning, social, and entertainment activities are beneficial, whereas the associations of internet-based work remain inconclusive. Economic interactions such as online shopping and digital transfers are positively associated with well-being, while increased reliance on virtual communication may reduce face-to-face interactions, weakening emotional connections.ConclusionThe “Broadband China” strategy played a significant role in promoting internet use among the country's older adults. While our results suggest many positive associations between internet use and mental well-being, increased internet usage is also associated with decreased face-to-face contact frequency and lower contact quality between older adults and their children.Policy ImplicationDigital inclusion policies should not only expand access but also promote balanced internet use while encouraging meaningful offline interactions between older adults and their families.

The existing literature is oriented toward examining how Internet access and use may affect social capital. The role of social capital in narrowing the digital divides has been frequently mentioned but few studies have empirically examined how various types of social capital may affect people's access and use of the Internet. Drawing on a two-wave national panel data set, this article aims to fill this gap. Results demonstrate that social capital facilitates Internet access and use. In particular resource-rich bonding social capital helps overcome the digital divides in access, general use, and online communication. Before the Internet can revitalize social capital, there must be the right social capital in place to close the digital divides. Highlighting the relationship between social connectivity and digital connectivity, the findings have important implications for policymakers and practitioners.