1,720,970 research outputs found
AMCIS: The Impact of Smartphone Usage on Health and Work
No full textIn the digital era, smartphones have become an integral part of our lives – much like an appendage. We have come to rely on these devices for communication, information, entertainment, and so much more. However, over-reliance on smartphones can have dire consequences. Prior research has found that excessive mobile phone use is associated with adverse consequences in different spheres of life (Nehra et al. 2012). Such adverse effects can extend to family, personal, and professional life (Zheng and Lee 2016). Younger people have even been found to prioritize using their phones over seeing loved ones (Die Presse 2018). Understanding why individuals use smartphones is relatively straightforward, but there is still not enough research into how individuals use smartphones and their effects on health and work. This research idea stemmed from studies that link information system (IS) usage to habit development (Limayem et al. 2007). In this case, the information system of interest is a smartphone, and we extend beyond habit development into health outcomes such as anxiety, stress, and addiction, as well as work outcomes such as productivity and efficiency. We aim to examine whether there is a relationship between how people use their smartphones and their health and work outcomes. When considering IS usage, three types of data exist – actual usage data, surveys, and interviews. While many studies, including Limayem et al. 2007, use surveys and interviews to analyze IS usage, we propose including additional actual usage data. Actual data in terms of smartphone usage can be extracted from activity-tracking smartphone applications. We can further expand the survey from smartphone usage items to include items that measure health and work outcomes. With these three types of data at hand, we can assess two things: (1) the impact of smartphone use (both actual and perceived) on health and work outcomes and (2) the difference between actual and perceived smartphone use. Furthermore, the actual usage data obtained from activity-tracking smartphone applications can uncover patterns of smartphone activity behavior that may be associated with stress, anxiety, or other outcomes. For example, patterns of phone checking in the absence of notifications or the sequence of actions performed by a subject may help us differentiate constructive from destructive behavior. Incorporating a survey also allows us to collect demographic data, which we can use to determine whether there are significant generational or gender differences in smartphone usage and their corresponding effects on health and work. To start examining possible outcomes linked to smartphone usage, we aim to undertake a pilot study
Understanding Long Covid-19 Patterns in Pediatric Patients using Network Analytics
No full textCOVID-19 has had a long-term impact on the quality of life, work, and society by manifesting in the form of Long COVID. Long COVID is relatively less understood condition due to its recency. The challenge is even greater among pediatric population, which represents 19% of overall Long COVID cases, due to lack of research and ample data. We study three main research questions in the study. First, what are the most frequently occurring chronic conditions among pediatric patients suffering from Long COVID at different time periods? Second, what are the most frequent non-chronic conditions among pediatric patients suffering from long-Covid across different age segments? Third, what are various clusters of chronic and non-chronic conditions that exist among pediatric patients diagnosed with Long-COVID.
Using N3C (National COVID Cohort Collaboration) data, we analyze health records of ~500K pediatric patients suffering from long COVID across 72 different sites. We apply network analytics approaches to model various chronic and non-chronic conditions that pre-exists in patients diagnosed with Long COVID. In the first part, we model two network types to capture the chronic and non-chronic diseases in pre and long Covid. In the second part, created bipartite graphs and its projections to generate network clusters of pre-existing diseases and coexisting disease network for Long COVID. We then applied two community detection algorithms, Louvain and Leiden algorithms, on these projections to identify clustering patterns of diseases. We analyzed and interpreted top clusters and observed a high dominance on the conditions related to the pregnancy, neoplasm(cancer), infectious, parasitic diseases and other categories. To develop insights into co-existing non-chronic conditions, we segmented the data across three pediatric age groups (0-4, 5-11, 12-17 years).
Our findings suggest that Long COVID co-exists with four highly frequent chronic conditions, namely, asthma, anxiety, obesity, and lipoprotein metabolism disorders. For all pediatric patients suffering from Long COVID, we found five dominant non-chronic co-existing conditions: acute upper respiratory infections, fever, Acute pharyngitis, deficit hyperactivity disorders, and cough. However, we observed some unique conditions when segmented across different age groups. For example, sleep disorders and severe stress were dominant across 11-17 age group. Using Louvain Community detection algorithm, we identified five key clusters. For example, cluster one (approx. 14% of data) had higher levels of teen pregnancies, infectious or parasitic diseases, and relatively lower levels of mental or behavioral disorders while cluster two (approx. 5.5%) had higher instances of neoplasms, and infectious or parasitic diseases. Our findings have important implications for pediatric care providers and researchers. Using network analytic approaches, we identified various clusters of chronic and non-chronic conditions that exist with Long COVID diagnosis among pediatric population. Such an understanding could provide early insights into the nature of pediatric patients who are likely to develop Long COVID from COVID-19
Health Analytics Lead to More Questions: A Comorbidity Lens Approach
No full textAs we amass more data, we have an opportunity to analyze a pseudo-population to better understand differences in health across groups. For example, comorbidity is a medical condition when a patient develops more than one disease simultaneously. The way patients belonging to different population groups develop comorbidities can have a major impact on their health outcomes. Therefore, there is a strong need to know these differences in comorbidities across population groups. In this study, we apply the grounded theory methodology lens to compare the comorbidities across population groups. First, we create a comprehensive network for each population group and then compare their structural properties. This leads to developing multiple research questions that need to be explored in the future research. The interesting findings and theoretical implications are discussed
Identifying Mortality Related Cliques in a Comorbidity Network
No full textA trap is defined as a situation where the entities in that trap are highly likely to experience an outcome. For example, a customer who has bought a certain group of items may be designated a high potential or high risk customer on some other attribute, going beyond the market basket analysis. The purpose of this paper is to detect traps related to a problem outcome by adapting the clique property of a network. We present a heuristics based method to first develop a latent network from the transactional data and then identify outcome related traps in the form of cliques. The method is demonstrated to detect mortality related traps of diseases in patients. We applied a network approach to create relationship between diagnoses and then used the clique property to identify high-risk traps of diagnoses. Using half of the patient records, the algorithm identified mortality related cliques in the network where the mortality rate in the patients diagnosed with all diseases is significantly higher than the rate in patients without all diagnoses in a clique. We validated the results on the other half of the patient records. The presence of the clique diagnoses in the patients can help physicians take preemptive treatment decisions to avoid letting a patient “fall into a trap” of the multiple diseases. The methodology can also be applied in other problems to find contextual traps
Exploring the role of online visual information in pre- and post- tourism experiences
No full textVisual information significantly influences tourists' pre and post tourism experiences; it creates a vivid stimulation of the experiences and can provoke the tourists' imagination to better convey the favorable information. Considering the role visual information plays, it has received much attention in business and academics, both from the perception of tourists and organizations. It is an important data channel to understand tourism experiences, travel patterns, and individual tourist's preferences. The development of information technology enables researchers to comprehend massive-scale structured and unstructured online data. It provides an easy way to explore how the tourists visualize the destinations and engage with the tourism activities.
The objective of this dissertation is revolved around visual information and its implication in the tourism field, along with other online displayed information, explores how they make an impact on post-tourism evaluation and pre-tourism decisions. In order to achieve this objective, this dissertation takes the form of three independent studies hoping to uncover the pictorial content relevance in the tourism field from three aspects:
1. A comprehensive systematic review of existing literature on visual information analysis in the hospitality and tourism field.
2. An experimental study on understanding pre-tourism experience decisions, via Search Engine Result Page (SERP) displays.
3. 2. A mixed-method approach aim to investigate post-experience evaluation on P2P gastronomy tourism, by factoring in web-based attributes, including web photos.
The following section breaks down and designates the highlights in each study.
Study 1 is a systematic literature review paper taken a text analytic approach that aims to review, analyze, and synthesize visual content analysis in tourism studies adapting big data analytic methods underlying machine learning. The literature search was conducted through two channels by keywords: Web of Science (WOS) and Scimago Journal & Country Rank (SJR) listed 124 tourism, leisure, and hospitality journals. In total, 67 papers were identified and considered after criteria filtering. To enhance the understanding of selected papers, first, a general description is provided with a distribution of research articles by journal fields and publication time. Second, a citation network was conducted to identify the connections between papers. Third, through text clustering, seven latent topics were discussed in detail in the results. The gap addressed and identified in study 1, study 2, and study 3 built on the aforementioned gaps and aims to amplify the significant role visual information plays in tourism behavior.
Study 2 aims to investigate information cues displayed through Airbnb experience Search Engine Result Page (SERP) and how it influences potential tourists' decision-making. In order to expand and explain tourists' information processing and identify the mechanism of displayed information cues on Airbnb at discrete levels, an experimental design was adopted. As such, to be able to assess the effects in a valid way, this study simulates Airbnb SERP and segmented information cues based on web collected data using the principle of math quartile for price, ratings, reviews, and time spent, and thematic categories for promotional photos, and promotional cues, including scarcity message, conformity message, and credibility cues, to test tourists' selection. All the thematic categories were decided from the text mining results. In addition, based on the selection of the results, tourists share similar selection characteristics to form a category, the market segmentations are formed based on the shared characteristics in selection. The incorporation of host credibility has never been explored in the context of P2P marketplace through an experimental design, the result from conjoint analysis shows the most important attribute is the host credibility, followed by images and price during consumers decision making. Theoretically, it employed signaling theory and heuristic systematic model in explaining how tourists’ processing online information. It successfully bridges the gap by introducing promotional photos and host credibility information cues in an experimental study. Methodologically, the text mining approach was taken in information cues segmentation, which serves as a solid foundation for conjoint analysis survey design. The survey-based conjoint analysis further identified important information cues to each level and provided insights into each market segment. The findings provide substantial practical implications for marketers, web information displays, and tourists' online information filtering, processing, and selection.
Study 3 selected gastronomy tourism, as a unique and growing tourism segmentation, aims to explore post-experience evaluation through Airbnb platform. This study adopted Satisfaction Disconfirmation, and the main contribution of this study is that it has taken a mixed-method approach in understanding gastronomy experiences post evaluation of the P2P platform (i.e., Airbnb). This research focuses on three areas of findings to explore post-evaluation and factors that might be influential to post-evaluation. First, this study examined 196,265 online reviews and used topic modeling to identify the major themes. The results have shown that LDA effectively identified six topics across all the reviews, which are (1) food city tour, (2) social interactions, (3) host, (4) local food culture, (5) beverage appreciation, (6) hands-on learning, ranked by topic importance. In addition to topic modeling, the study further performed sentiment analysis, aim to generate different sentiments reviews count for gastronomy experiences. Second, through Google cloud vision API, 1,331 photos were analyzed, and gained image insights through three aspects: colorfulness, face detection, object detection. Colorfulness was further integrated into OLS regression. Third, OLS regression explored the factors that have significant impacts on gastronomy experiences ratings, with a consideration of review sentiments, image colorfulness, host credibility indicator (i.e., length of experience), and neutralized review count (review count/length of experiences). Theoretically, it employed expectation disconfirmation theory, and this is one of the first studies exploring tourism experiences provided through P2P platforms and evaluating the impact of both textual data and image data on gastronomy experiences. It successfully combined big data analytics (e.g., topic modeling, sentiment analysis, image analysis) and traditional methods (i.e., OLS regression). Practically, for the hosts and service providers, the study findings provide insights into aspects that gastronomy tourists’ care more about, and factors that influence overall satisfactions of post evaluation
Using AI Techniques to Understand Text in Different Contexts: Misinformation, COVID-19, and Mental Health
No full textThis dissertation work consists of three studies. This work uses a multi-pronged approach that draws upon various forms of online textual data using Partial Least Squares Structural Equation Modeling (PLS-SEM) and different Artificial Intelligence (AI) techniques such as text analytics, natural language processing (NLP), network analytics, and deep learning approaches to understand the misinformation research, dynamics of COVID-19 news topics on Facebook, and how humans evaluate AI-generated and human-generated reviews in mental health apps.
Study 1 provided a systematic text-analytic literature review on misinformation research. First, this survey study presented a framework to explain the creation and spread of misinformation within the AI context. Second, we synthesized four thematic dimensions where recent research on misinformation has been evolving: (1) spread of misinformation, (2) impacts of misinformation, (3) misinformation detection, and (4) mitigation of misinformation. This effort serves to provide guidance for IS and Analytics researchers in pursuing this important line of research on developing improved models and analytics approaches that help mitigate the impact of misinformation.
Study 2 examined the evolution of news topics during the COVID-19 pandemic across progressively different timeframes based on seven phases of the pandemic cycle. Guided LDA analysis of 30 million Facebook English posts collected over all the COVID-19 phases helped us identify key terms across each phase. Some topics (terms) were identified as among the most relevant topics in all phases. These key topics can be grouped into four categories: the number of new and death cases (e.g., “coronavirus case”, “test positive”, “new case”, and “positive coronavirus”, prevention strategies (e.g., “close contact”, “social distance”, “wash hand”, “stay home”, “home test”, and “test kit”), COVID-19 vaccines (e.g., “coronavirus vaccine”, “coronavirus vaccination”, “second dose”, and “fully vaccinate”.), and variants (e.g., “delta variant” and “omicron variant”). Thematic analysis revealed ten dominant themes among different COVID-19 phases. Interestingly, we found that even though some themes (e.g., mitigation and prevention strategies) are prevalent across all phases, the topics within these themes evolve as the COVID-19 cycle progresses. Furthermore, the theme “Severity of COVID-19” became dominant during the middle stages of the pandemic when COVID-19 showed a downward trend. Counterintuitively, text network analytics uncovered that topics during the early stages of the pandemic were more connected than the phases when COVID-19 showed an upward trend, indicating that social media topics cover a wider range of topics. Findings from this study could be used to develop better information dissemination portals and content for different pandemic cycles. This study also guides various responsible agencies on synchronizing messaging based on pandemic phases and future pandemic planning.
Study 3 investigated how app users are stimulated by the linguistic cues from online reviews and how they process reviews through different paths of internal states, such as perceived cognitive effort and perceived persuasion motives, which affect perceived review credibility. This study confirmed that consumers’ internal states have different impacts on the perceived review credibility, depending on whether the reviews are AI-generated or human-generated. In AI-generated reviews, internal states are affected by Emotion only. However, regarding human-generated reviews, internal states are affected by Complexity, Emotion, and Uncertainty. These findings open up opportunities for future research to incorporate more features of AI-generated fake reviews
Analysis of a Sequence of Events in Healthcare
No full textHealthcare industry generates streams of data in different problem domains. Analysis of such data requires stream analytics tools and techniques to generate useful insights. Stream analytics involve analysis of time variant events. The specific patterns in the events can indicate some imminent outcomes such as state of a heart, etc. Therefore, novel ways to find specific patterns in the events generated by multiple sources are required. A key requirement for applying any such method is data preparation and organization to enable such analysis. In this paper, we extend the CRISP-DM process to include data preparation approaches for sequence mining. We present progression analysis, an approach for converting multidimensional time variant streams of health records in a form to be able to detect useful sequential signals. To illustrate the process, we use patient health history stored in an Electronic Medical Record system (EMR) and present a healthcare application to compare progression of diseases over time between patients diagnosed with Tobacco Use Disorder (TUD) and non-tobacco users. Interestingly, many diseases follow the same path for TUD and non-TUD patients. Finally, the generalizability of the progression analysis is discussed
Teaching Social Network Analytics via Virtual Reality: Demonstrating a Novel VR Application
No full textPractitioners and researchers have extensively utilized network analysis to explore interactions among various elements. Traditionally, networks are mathematically represented using matrices. Subsequently, two-dimensional visualization tools became popular for analyzing these networks. In educational settings, the current approach involves teaching networks through matrices and 2-D visualizations. This paper introduces an innovative approach to teaching network analysis using virtual reality (VR) in which the user is a part of the network. By leveraging VR, students can gain a deeper understanding of network concepts through immersive interaction with the network by being a part of it. Specifically, the application will demonstrate how the concepts of a social network can be illustrated to potential learners of social network analytics
Domino Effects of AI: Spillovers from New App Launches on Developer Portfolio
No full textDisruptive innovations often transform digital platforms by reshaping how value is created and consumed, altering dynamics across the entire ecosystem. Among these, AI has emerged as a fast-moving and transformative force. Mobile applications serve as a key channel for delivering AI-powered services, making them crucial for studying AI’s broader impacts. This study explores how launching a new AI app affects the demand for a developer’s existing apps and how this relationship depends on the AI composition within the developer’s portfolio. Leveraging data mining, we constructed a unique six-month biweekly panel dataset from Apple’s App Store. Two-way fixed effects regression models reveal that new AI app launches increase demand for existing apps, especially when the developer’s portfolio is primarily non-AI. However, this effect weakens as the portfolio becomes more AI-heavy. These findings contribute to the demand-spillover and disruption literature and offer practical insights for managing AI integration in digital ecosystems
- …
