Predicting aspect-based sentiment using deep learning and information visualization: The impact of COVID-19 on the airline industry

2022 | Information & Management | Citations: 0

Authors: Chang, Yung-Chun; Ku, Chih-Hao; Nguyen, Duy-Duc Le

Abstract: This study investigates customer satisfaction through aspect-level sentiment ana ... Expand

Abstract: This study investigates customer satisfaction through aspect-level sentiment analysis and visual analytics. We collected and examined the flight reviews on TripAdvisor from January 2016 to August 2020 to gauge the impact of COVID-19 on passenger travel sentiment in several aspects. Till now, information systems, management, and tourism research have paid little attention to the use of deep learning and word embedding techniques, such as bidirectional encoder representations from transformers, especially for aspect-level sentiment analysis. This paper aims to identify perceived aspect-based sentiments and predict unrated sentiments for various categories to address this research gap. Ultimately, this study complements existing sentiment analysis methods and extends the use of data-driven and visual analytics approaches to better understand customer satisfaction in the airline industry and within the context of the COVID-19. Our proposed method outperforms baseline comparisons and therefore contributes to the theoretical and managerial literature. Collapse

Topics: pandemic customer satisfaction visual analytics cloud computing data visualization

Methods: BERT deep learning natural language processing exploratory data analysis sentiment analysis

Theories: construal level theory appraisal theory

Write Like a Pro or an Amateur? Effect of Medical Language Formality

2021 | ACM Transactions on Management Information Systems | Citations: 4

Authors: Xie, Jiaheng; Zhang, Bin; Brown, Susan; Zeng, Daniel

Abstract: Past years have seen rising engagement among caregivers in online health communi ... Expand

Abstract: Past years have seen rising engagement among caregivers in online health communities. Although studies indicate that this caregiver-generated online health information benefits patients, how such information can be perceived easily and correctly remains unclear. This study aims to fill this gap by exploring mechanisms to improve the perceived helpfulness of online health information. We propose a multi-method framework, including a novel Medical-Enriched DEep Learning (MEDEL) feature extraction method, econometric analyses, and a randomized experiment. The results show that when the medical language of health information is informal, the senior care information is more helpful. Our findings provide a theoretical foundation to understand the influence of language formality on many other business communications. Our proposed multi-method approach can also be generalized to investigate research questions involving complex textual features. Forum sites could leverage our proposed approach to improve the helpfulness of online health information and user satisfaction. Collapse

Topics: healthcare data user satisfaction data quality knowledge base health information system

Methods: deep learning propensity score matching experiment experimental group word embedding

Artificial Intelligence and Visual Analytics: A Deep-Learning Approach to Analyze Hotel Reviews & Responses

2019 | HICSS | Citations: 0

Authors: Ku, Chih-Hao; Tech, Lawrence; Chang, Yung-Chun; Wang, Yichung; Chen, Chien-Hung

Abstract: With a growing number of online reviews, consumers often rely on these reviews t ... Expand

Abstract: With a growing number of online reviews, consumers often rely on these reviews to make purchase decisions. However, little is known about managerial responses to online hotel reviews. This paper reports on a framework to integrate visual analytics and machine learning techniques to investigate whether hotel managers respond to positive and negative reviews differently and how to use a deep-learning approach to prioritize responses. In this study, forty 4- and 5-star hotels in London with 91,051 reviews and 70,397 responses were collected and analyzed. Visual analyses and machine learning were conducted. The results indicate most hotels (72.5%) showing no preference to respond to positive and negative reviews. Our proposed deeplearning approach outperformed existing algorithms to prioritize responses. Collapse

Topics: visual analytics online review decision making Python user satisfaction

Methods: product review exploratory data analysis deep learning cluster analysis machine learning