How product review voting is influenced by existing votes, consumer involvement, review valence, and review diagnosticity

2023 | Decision Support Systems | Citations: 0

Authors: Tseng, Shih-Lun; Lu, Shuya; Weathers, Danny; Grover, Varun

Abstract: Websites that allow consumers to post online reviews often allow others to signa ... Expand

Abstract: Websites that allow consumers to post online reviews often allow others to signal the value of the reviews by voting for them as being helpful (i.e., an upvote) or unhelpful (i.e., a downvote). Although a review's content should drive votes, herd behavior may also be operative if consumers discount their own judgments of the review based on signals others provide. Using a field study of online product reviews from Amazon, a longitudinal experiment, and two cross-sectional experiments, we find evidence for herd behavior: consumers' voting de­ cisions and intentions are influenced by others' votes. However, this effect is complex with multiple moderating factors, including consumer involvement, review valence, and review diagnosticity. We discuss the theoretical and practical implications of these findings for reducing the potential negative effects of herd behavior in the context of online reviews. Collapse

Topics: online review Reddit Twitter electronic commerce user-generated content

Methods: computational algorithm longitudinal research longitudinal experiment statistical hypothesis test parametric test

Theories: signaling theory

A taxonomy for deriving business insights from user-generated content

2023 | European Conference On Information Systems | Citations: 0

Authors: Weingart, Pablo; Wambsganss, Thiemo; Soellner, Matthias

Abstract: Deriving business insights from user-generated content (UGC) is a widely invest ... Expand

Abstract: Deriving business insights from user-generated content (UGC) is a widely investigated phenomenon in information systems (IS) research. Due to its unstructured nature and technical constraints, UGC is still underutilized as a data source in research and practice. Using recent advancements in machine learning research, especially large language models (LLMs), IS researchers can possibly derive these insights more effectively. To guide and further understand the usage of these techniques, we develop a taxonomy that provides an overview of business insights derived from UGC. The taxonomy helps both practitioners and researchers identify, design, compare and evaluate the use of UGC in this IS context. Finally, we showcase an LLM-supported demo application that derives novel business insights and apply the taxonomy to it. In doing so, we show exemplary how LLMs can be used to develop new or extend existing natural language processing (NLP) applications in the realm of IS. Collapse

Topics: database system social network online review social media user-generated content

Methods: natural language processing structured literature research qualitative interview deep learning NLP application

Decide Now or Later: Making Sense of Incoherence Across Online Reviews

2023 | Information Systems Research | Citations: 0

Authors: Yin, Dezhi; de Vreede, Triparna; Steele, Logan M.; de Vreede, Gert-Jan

Abstract: Mixed or inconsistent opinions are commonplace in online reviews. Prior research ... Expand

Abstract: Mixed or inconsistent opinions are commonplace in online reviews. Prior research shows that review inconsistency has different effects: its product-level manifestation in the form of inconsistent product ratings is associated with poorer sales, but its review-level manifestation in the form of two-sided arguments is associated with greater helpfulness and credibility evaluations of the review. In practice, consumers rarely consult all reviews or just a single review before they make purchase decisions. Instead, they often read a set of featured reviews (i.e., a review set) and some additional reviews if needed. Focusing on inconsistency in a review set, we introduce a new type of inconsistency across reviews that does not exist within a single review or at the level of product ratings: cross-review incoherence, which refers to disagreement among reviewers about specific product features. Based on cognitive dissonance theory, we explore how and when cross-review incoherence influences helpfulness and credibility judgments of the review set, revealing that consumers’ reactions to such incoherence might operate differently beyond an individual review. In addition, we examine how and why consumers’ judgments of a review set influence their purchase deferral—that is, the likelihood of making a buy-or-not-buy decision immediately after consulting the top reviews or deferring it until after obtaining more information. Two laboratory experiments demonstrate that cross-review incoherence increases purchase deferral via more negative evaluations of the review set that, in turn, reduce attitude certainty. In addition, the negative effect of cross-review incoherence on review set evaluations is weaker when the reviewers provide more contextual information behind their opinions. These findings deepen our understanding of inconsistency across multiple pieces of information; reveal the consequences of review evaluations during an understudied stage of consumers’ decision-making process; and provide critical implications for review platforms, companies, and reviewers. History: Paul Pavlou, Senior Editor; David (Jingjun) Xu, Associate Editor. Supplemental Material: The online appendix is available at https://doi.org/10.1287/isre.2022.1150. Collapse

Topics: decision making online review user-generated content data relevance enterprise information system

Methods: experiment chi squared test laboratory experiment conceptual modelling theory development

Theories: cognitive dissonance theory

Extracting Actionable Insights from Text Data: A Stable Topic Model Approach

2023 | Management Information Systems Quarterly | Citations: 0

Authors: Yang, Yi; Subramanyam, Ramanath

Abstract: Topic models are becoming a frequently employed tool in the empirical methods r ... Expand

Abstract: Topic models are becoming a frequently employed tool in the empirical methods repertoire of information systems and management scholars. Given textual corpora, such as consumer reviews and online discussion forums, researchers and business practitioners often use topic modeling to either explore data in an unsupervised fashion or generate variables of interest for subsequent econometric analysis. However, one important concern stems from the fact that topic models can be notorious for their instability, i.e., the generated results could be inconsistent and irreproducible at different times, even on the same dataset. Therefore, researchers might arrive at potentially unreliable results regarding the theoretical relationships that they are testing or developing. In this paper, we attempt to highlight this problem and suggest a potential approach to addressing it. First, we empirically define and evaluate the stability problem of topic models using four textual datasets. Next, to alleviate the problem and with the goal of extracting actionable insights from textual data, we propose a new method, Stable LDA, which incorporates topical word clusters into the topic model to steer the model inference toward consistent results. We show that the proposed Stable LDA approach can significantly improve model stability while maintaining or even improving the topic model quality. Further, employing two case studies related to an online knowledge community and online consumer reviews, we demonstrate that the variables generated from Stable LDA can lead to more consistent estimations in econometric analyses. We believe that our work can further enhance management scholars' collective toolkit to analyze ever-growing textual data. Collapse

Topics: online review user-generated content open source usability social media

Methods: topic model experiment word embedding Gibbs sampling computational algorithm

CEO’s Communication Styles and their Effect on Organizational Performance

2022 | Americas Conference on Information Systems | Citations: 0

Authors: Bojja, Giridhar Reddy; Liu, Jun; Shan, Ronghua; Wang, Yong

Abstract: Several studies have highlighted the importance of leadership communication in ... Expand

Abstract: Several studies have highlighted the importance of leadership communication in establishing stakeholder alignment, implementing strategy, and achieving superior organizational performance. In this way, this study investigates the impact of CEO's communication styles on organizational performance using textmining approach on CEOs tweets from social media. This study aims to examine the relationship between CEO communication styles and organizational performance. The contribution of our study is three-folded: 1) From a methodological standpoint, we present a model to establish a relationship between CEO communication styles on social media and firm performance. Additionally, we apply text mining to identify communication styles of CEOs. 2) From a performance management, we evaluate organizational performance in three types: Operational, Financial, and Reputational. 3) From a management practice and policy perspective, our study results will help organizations evaluate the CEO candidates from a communication style standpoint. IntroductionThe importance of CEOs has been well recognized in strategic management literature. CEOs represent the company as the highest-ranking corporate spokesman responsible for communicating the company's goals (Chen and Chung 2016). A successful CEO not only strengthens public identification with the business and generates a strong corporate reputation, but also builds beneficial relationships with key stakeholders to achieve financial benefits (Tsai and Men 2017). According to upper echelons theory, a CEOs observable and psychological characteristics influence his or her decisions, and thus company performance. CEO communication is one of the major observable and psychological characteristic that can help understand CEO behavior. With the growing need for effective corporate governance and authenticity fueled by social media, the public has tended to recognize greater access to the views and vision of corporate executives through more open interactions. CEOs' public personas have shifted to one that is more personable, sociable, and open to the public than ever before. As a leader of the organization, CEOs have greater responsibility to engage both internal and external stakeholders. Social media is a powerful communication medium, with widespread influence all over the world. CEOs' can leverage social media platforms such as Twitter, to interact and engage with various stakeholders. As a key psychological characteristic, CEO communication has been considered as one of the key managerial cognitive abilities that achieve superior business performance (Bakker-Pieper and de Vries 2013). Thus, it is essential to understand the impact of CEO communication styles on organizational performance.Previous studies on CEO behavior majority looked at the emotionality (Rajah et al. 2011) and personality (Andersen 2006) aspect of CEOs. As per as per (Bakker-Pieper and de Vries 2013) the communication styles are more relevant and were strongly linked to the leader outcomes than emotionality and personality traits. Communication styles are closer to actual leadership styles as they are more flexible and subject to CEO Communication and Organizational Performance Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022 2 change as per leader situation and may be exhibited as a mix of more than one style over time. From a comprehensive view, communication comprises multiple dimensions and all dimensions take effect simultaneously when a CEO makes decisions. So, we adopt (Schulz von Thun 1983) communication theory, which is the widely accepted framework for measuring communication types, to capture every aspect of CEO communication styles. Research in this area requires studies to investigate relationship between CEO communication styles and organizational performance. This problem represents a pertinent gap in Upper Echelons literature. Therefore, the purpose of this study is to present a model to establish a relationship between CEO communication styles on social media and firm performance. Additionally, we apply text mining, which is not commonly used in to identify communication styles of CEO in the literature. The communication styles being the variables of interest in the study are primarily drawn from (Stajner et al. 2021) include Self-Revealing, Action-Seeking, Fact-Oriented, and Information-Seeking. They may also be referred to as psycholinguistic cues representing the style or state respectively (Stajner et al. 2021). Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022 4and set deadlines for others to follow. Usually, they back it up with some form of payoff. Action-seeking or task-oriented leadership communication has been acknowledged as particularly useful in emergency situations, or at times, when the company must be going through a crisis. In such cases, action seeking communication is synonymous with providing a definitive direction, and thus is likely to improve specific efficiency and productivity outcomes, falling mainly within an operational domain (Groves 2007). Hence, Hypothesis 3: There is a significant impact of a CEO's action-seeking style on organizational performance variables.Information-seeking (I-S) communication style is one, wherein direct or indirect questions are sought from others (Stajner et al. 2021). In social-media communication terms, the information-seeking communication style involves engaging others by posing questions or asking for advice (Symanto, 2022). Within the CEO communication sphere, information-seeking communication largely connotes advice-seeking communication (Ma et al. 2020). Such communication behavior has been found to have important effects on business and operational performance aspects like strategic decision making, management effectiveness, entrepreneurial orientation, innovation, and financial performance. Such CEO communication also likely affects relative competitive firm performance (Heyden et al. 2013). Hence, Hypothesis 4: There is a significant impact of a CEO's information-seeking style on organizational performance variables. Conclusion and LimitationsThis study aimed to find the relationship between CEO communication styles and organizational performance. This research answers a critical question in terms of organizational strategy literature, what dimensions of CEO communication styles affect organizational performance and in which direction. The study contributes in multiple ways: (1). We present a model to establish a relationship between CEO communication styles on social media and firm performance. Additionally, we apply text mining, which is not commonly used in to identify communication styles of CEO in the literature. AI-based text mining strategy based on social media data to quantify the CEO's communication styles (2). Theoretically, we contribute to strategic management and upper echelon theory. By far, we are the first to examine CEO communication styles across three types of organizational performance: operational, financial, and reputational (3). In practice, our findings provide a foundation for organizations to assess executive candidates from a communication standpoint.Our results do have some limitations: 1) We are restricted to a small dataset since only a few CEOs were active on social media, in our case, Twitter. 2) We only used a single social media platform, "Twitter", to crawl the CEO's textual data. Crawling from multiple social media sites such as Facebook, personal blogs will provide holistic data, thus providing rich insights. 3) Our results provide an overall effect of CEO communication styles on organizational performance. Future studies may identify the potential mediating and moderating paths to provide better organization performance. Collapse

Topics: organizational productivity Twitter social media application programming interface website

Methods: natural language processing company material Compustat deep learning cross sectional research

Theories: communication theory

Space Norms for Constructing Quality Reviews on Online Consumer Review Sites

2022 | Information Systems Research | Citations: 0

Authors: Hou, Jinghui (Jove); Ma, Xiao

Abstract: High-quality online consumer reviews are valuable to e-commerce stakeholders. Ex ... Expand

Abstract: High-quality online consumer reviews are valuable to e-commerce stakeholders. Extant research mainly focuses on what drives consumers to submit reviews, whereas little has been done to specifically theorize why review writers construct high-quality reviews versus post poor review content. To advance current theorizing on the behavior of writing quality reviews, we propose the concept of space norms, a distinct instantiation of social norms inferred from locally observing how others write reviews on the local digital space (i.e., the review site). Specifically, we theorize that the focal behavior is largely directed by space norms; space norms beget a sense of how one should do and induce conformity in the given social space. We tested the theory of space norms in a series of empirical studies, and the findings provide converging evidence. Study 1 used experimental methods and demonstrates—with large effect sizes—that review writers tend to act in accordance with the space norm of the local site. Post hoc text analyses rule out an alternative explanation and corroborate the findings. Study 2 used field data collected from two e-commerce websites to further triangulate the space norms proposition. Another post hoc study confirms the robustness of the results. This paper contributes to information systems research by developing and empirically validating a theory of space norms that pinpoints the role of local digital spaces in constructing the focal normative phenomenon. It also lends insights to digital platforms and the information technology practitioners. Managers looking to elicit quality reviewing behavior should devise the local digital space to nurture the space norms for the desired behavior. Collapse

Topics: social presence online review social norm website electronic commerce

Methods: theory development natural language processing computational algorithm statistical hypothesis test design artifact

Fraudulent review detection model focusing on emotional expressions and explicit aspects: investigating the potential of feature engineering

2022 | Decision Support Systems | Citations: 5

Authors: Kumar, Ajay; Gopal, Ram D.; Shankar, Ravi; Tan, Kim Hua

Abstract: Reading customer reviews before purchasing items online has become a common prac ... Expand

Abstract: Reading customer reviews before purchasing items online has become a common practice; however, some companies use machine learning (ML) algorithms to generate false reviews in order to create positive brand images of their own products and negative images of competitors' offerings. Existing techniques use review content to identify fraudulent reviewers; however, spammers become more intelligent, started to learn from their mistakes, and changed their tactics in order to avoid detection techniques. Thus, investigating fraudulent ac­ counts' behaviour of generating fake negative or positive reviews for competitors or themselves and the necessity of ML classifiers to identify fraudulent reviews, is more important than ever. In this research, we present a novel feature engineering approach in which we (1) extract several “review-centric” and “reviewer-centric” features from a dataset; (2) combine the cumulative effects of features distributions into a unified model that represents overall behavior of the fraudulent reviewers; (3) investigate the role of effective data pre-processing to improve detection accuracy; and (4) develop a probabilistic approach to detect fraudulent reviewers by learning a novel M-SMOTE model over a derived balanced dataset and feature distributions, which outperforms other ML models. Our study contributes to the literature on digital platforms and fraudulent review detection with significant managerial and theoretical implications through these novel findings. Collapse

Topics: online review electronic commerce website marketing management logistics management

Methods: feature engineering machine learning computational algorithm XGBoost product review

ASPECT-BASED SENTIMENT ANALYSIS FOR UNIVERSITY TEACHING ANALYTICS

2022 | European Conference On Information Systems | Citations: 0

Authors: Schurig, Tim; Zambach, Sine; Mukkamala, Raghava Rao; Petry, Malte

Abstract: Aspect-based sentiment analysis (ABSA) is a natural language processing method ... Expand

Abstract: Aspect-based sentiment analysis (ABSA) is a natural language processing method to analyze sentiments from large amounts of unstructured text in a much more fine-grained manner at the aspect level. In this research work, we apply it to analyze open text replies from surveys regarding online teaching. Like most other educational institutions, Copenhagen Business School (CBS) had to shift to online teaching from one day to the next. Using ABSA, we investigated the impact of this forced online learning experiment on teaching quality in the spring semester of 2020. Our findings reveal that students disliked online teaching due to insufficient information and unadjusted teaching methods. However, students liked its flexibility and possibility to learn at an individual pace. We show that ABSA can extract valuable information in an easily interpretable manner to support teaching and learning processes. Finally, our findings show that ABSA is a valuable tool to analyze unstructured text quantitatively. Collapse

Topics: online learning system Python decision making

Methods: survey literature study qualitative interview natural language processing NLTK

Measuring Product Type and Purchase Uncertainty with Online Product Ratings: A Theoretical Model and Empirical Application

2021 | Information Systems Research | Citations: 7

Authors: Chen, Peiyu; Hitt, Lorin M.; Hong, Yili; Wu, Shinyi

Abstract: Building on the distinction between search and experience goods, as well as vert ... Expand

Abstract: Building on the distinction between search and experience goods, as well as vertical and horizontal differentiation, we propose a set of theory-grounded, data-driven measures that allow us to measure not only product type (search vs. experience, horizontal vs. vertical differentiation) but also sources of uncertainty and to what extent consumer reviews help resolve uncertainty. The proposed measures have two advantages over prior methods: (1) unlike prior categorization schemes that classified goods as either search or experience goods, our measure is continuous, allowing us to rank-order the degree of search versus experience and horizontal versus vertical differentiation among products or categories. (2) Our approach is easier to implement than prior methods, because it relies solely on consumer ratings information (as opposed to expert judgment) and can be employed at multiple levels (attributes, products, or product categories). We illustrate empirical applications of our proposed measures using product rating data from Amazon.com. Our data-driven measures reveal the relative importance of fit in driving product utility and the importance of search for determining fit for each product category at Amazon. Our results also show that, while ratings based on verified purchasers are informative of objective product values, the current Amazon review system appears to have limited ability to resolve fit uncertainty. Our method and findings could facilitate further research on product review systems and enable quantitative measurement of product positioning to support marketing strategy for retailers and manufacturers, covering an expanded group of products. Collapse

Topics: experience good horizontal differentiation product differentiation online review marketing management

Methods: longitudinal research Amazon review descriptive statistic product review time series analysis

Intention-based Deep Learning Approach for Detecting Online Fake News

2021 | International Conference on Information Systems | Citations: 0

Authors: Lee, Kyuhan; Ram, Sudha

Abstract: One effective approach to fight fake news is to automatically filter it out usi ... Expand

Abstract: One effective approach to fight fake news is to automatically filter it out using computational approaches. However, current approaches have neglected to identify the intention behind posting fake news leading to errors in flagging fake news. In this study, following the design science approach, we propose a novel deep-learning framework for detecting online fake news by incorporating theories of deceptive intention. Specifically, we first develop a transfer-learning model that identifies deceptive intention reflected in text and apply it to distinguish two subclasses of fake news: deceptive and non-deceptive fake news. Then, these two classes of fake news, along with an observed class of non-fake news (i.e., true news), are used to train deep bidirectional transformers whose goal is to determine news veracity. Our framework is empirically evaluated and benchmarked against cutting-edge deep learning models. Our analysis reveals that the models incorporating our deceptive-intention-based design significantly outperform state-of-theart baselines. Collapse

Topics: fake news online review application programming interface cloud computing personality

Methods: deep learning machine learning natural language processing design science part of speech tagging

Theories: cognitive complexity theory