Improved healthcare disaster decision-making utilizing information extraction from complementary social media data during the COVID-19 pandemic

2023 | Decision Support Systems | Citations: 0

Authors: Kellner, Domenic; Lowin, Maximilian; Hinz, Oliver

Abstract: Managing an extreme event like a healthcare disaster requires accurate informati ... Expand

Abstract: Managing an extreme event like a healthcare disaster requires accurate information about the event's circum­ stances to comprehend the full consequences of acting. However, information quality is rarely optimal since it takes time to determine the information of relevance. The COVID-19 pandemic showed that even official data sources are far from optimal since they suffer from reporting delays that slow decision-making. To support decision-makers with timely information, we utilize data from online social networks to propose an adaptable information extraction solution to create indices helping to forecast COVID-19 case numbers and hospitalization rates. We show that combining heterogeneous data sources like Twitter and Reddit can leverage these sources' inherent complementarity and yield better predictions than those using a single data source alone. We further show that the predictions run ahead of the official COVID-19 incidences by up to 14 days. Additionally, we highlight the importance of model adjustments whenever new information becomes available or the underlying data changes by observing distinct changes in the presence of specific symptoms on Reddit. Collapse

Topics: Twitter Reddit social media pandemic website

Methods: BERT information extraction computational algorithm autoregressive integrated moving average time series analysis

A Modular Social Sensing System for Personalized Orienteering in the COVID-19 Era

2023 | ACM Transactions on Management Information Systems | Citations: 0

Authors: Pilato, Giovanni; Persia, Fabio; Ge, Mouzhi; Chondrogiannis, Theodoros; D’Auria, Daniela

Abstract: Orienteering or itinerary planning algorithms in tourism are used to optimize tr ... Expand

Abstract: Orienteering or itinerary planning algorithms in tourism are used to optimize travel routes by considering user preference and other constraints, such as time budget or traffic conditions. For these algorithms, it is essential to explore the user preference to predict potential points of interest (POIs) or tourist routes. However, nowadays, user preference has been significantly affected by COVID-19, since health concern plays a key tradeoff role. For example, people may try to avoid crowdedness, even if there is a strong desire for social interaction. Thus, the orienteering or itinerary planning algorithms should optimize routes beyond user preference. Therefore, this article proposes a social sensing system that considers the tradeoff between user preference and various factors, such as crowdedness, personality, knowledge of COVID-19, POI features, and desire for socialization. The experiments are conducted on profiling user interests with a properly trained fastText neural network and a set of specialized Naïve Bayesian Classifiers based on the “Yelp!” dataset. Also, we demonstrate how to approach and integrate COVID-related factors via conversational agents. Furthermore, the proposed system is in a modular design and evaluated in a user study; thus, it can be efficiently adapted to different algorithms for COVID-19-aware itinerary planning. Collapse

Topics: user satisfaction social network personality conversational agent Twitter

Methods: computational algorithm fastText machine learning experiment descriptive statistic

Identifying Structural Holes for Sentiment Classification

2022 | Information Systems Frontiers | Citations: 0

Authors: Xie, Zheng; Liu, Guannan; Qu, Jinming; Wu, Junjie; Li, Hong

Abstract: The prevalence of online user-generated content has attracted great interest in ... Expand

Abstract: The prevalence of online user-generated content has attracted great interest in textual sentiment analysis, which provides a low-cost yet effective way to discern consumers and markets. A mainstream of sentiment analysis is to construct a classification model with Bag-of-Words (BoW) features, but the large vocabulary base and skewed distribution of term frequency consistently pose research challenges, which is made even worse by the limited valid sentiment labels. In light of this, in this paper, we propose a novel method called Structural Holes based Sentiment Classifier (SHSC) for BoW-based sentiment classification. The key to SHSC is to reinforce the classification contribution of semantically rich words with clear-cut sentiment polarity. To this end, a word co-occurrence network is carefully constructed to represent both high and low frequency words. The work to find classification-inefficient words is then transformed into the identification of so-called bridge nodes that occupy the positions of structural holes in the network. Two interesting measures, i.e., information advantage rank and control advantage weight, are then designed elaborately for this purpose, which are based on the proposed sentiment-label propagation and short-path computation algorithms, respectively. SHSC finally feeds this information as the key regularizers into a simple regression model to guide parametric learning. Extensive experiments on real-world text datasets demonstrate the advantage of our SHSC model over competitive benchmarks, particularly when sentiment labels are scarce. The effectiveness of uncovering structural holes for sentiment classification is also carefully verified with some robustness checks and demonstration cases. Collapse

Topics: logistics management microblog Twitter Python Wikipedia

Methods: experiment machine learning Student's t-test support vector machine word embedding

Theories: graph theory

Cross-Category Defect Discovery from Online Reviews: Supplementing Sentiment with Category-Specific Semantics

2022 | Information Systems Frontiers | Citations: 0

Authors: Zaman, Nohel; Goldberg, David M.; Gruss, Richard J.; Abrahams, Alan S.; Srisawas, Siriporn; Ractham, Peter; Şeref, Michelle M.H.

Abstract: Online reviews contain many vital insights for quality management, but the volum ... Expand

Abstract: Online reviews contain many vital insights for quality management, but the volume of content makes identifying defect-related discussion difficult. This paper critically assesses multiple approaches for detecting defect-related discussion, ranging from out-of-the-box sentiment analyses to supervised and unsupervised machine-learned defect terms. We examine reviews from 25 product and service categories to assess each method’s performance. We examine each approach across the broad cross-section of categories as well as when tailored to a singular category of study. Surprisingly, we found that negative sentiment was often a poor predictor of defect-related discussion. Terms generated with unsupervised topic modeling tended to correspond to generic product discussions rather than defect-related discussion. Supervised learning techniques outperformed the other text analytic techniques in our cross-category analysis, and they were especially effective when confined to a single category of study. Our work suggests a need for category-specific text analyses to take full advantage of consumer-driven quality intelligence. Collapse

Topics: online review experience good logistics management information retrieval user-generated content

Methods: machine learning sentiment analysis natural language processing supervised learning cross sectional research

FairPlay: Detecting and Deterring Online Customer Misbehavior

2021 | Information Systems Research | Citations: 4

Authors: Wu, Ji; Zheng, Zhiqiang (Eric); Zhao, J. Leon

Abstract: Customer misbehavior is a serious and pervasive problem in firm-sponsored social ... Expand

Abstract: Customer misbehavior is a serious and pervasive problem in firm-sponsored social media, yet prior studies provide limited insight into how firms should detect and manage it. To address this gap, we first develop a data science approach to detect customer misbehavior on social media and then devise intervention strategies to deter it. Specifically, we build on natural language processing and deep learning techniques to automatically detect customer misbehavior by mining customers’ social media activities in collaboration with a leading apparel firm. The results show that our algorithmic solution achieves superior performance, improving detection by 7%–9% compared with traditional methods. We then implement two types of intervention policies based on the focus theory of normative conduct that advocates the use of injunctive norms (i.e., a punishment policy) and descriptive norms (i.e., a common identity policy) to restrain customer misbehavior. We conduct field experiments with the firm to validate these policies. The experimental results indicate that punishment considerably reduces customer misbehavior in the short term, but this effect decays over time, whereas common identity has a smaller but more persistent effect on misbehavior reduction. In addition, punishing dysfunctional customers decreases their purchase frequency, whereas imposing a common identity increases it. Interestingly, our results show that combining the two policies effectively alleviates the detrimental effect of punishment, especially in the long run. We examine the heterogeneous treatment effect on novice and experienced customers. Finally, a follow-up field experiment reveals that the disclosure of the use of an artificial intelligence detector improves the effectiveness of the intervention strategies, and this effect is more pronounced for the punishment and combination strategies. Collapse

Topics: artificial intelligence social media social norm logistics management online learning system

Methods: field experiment computational algorithm experiment experimental group word embedding