Towards Hybrid Architectures: Integrating Large Language Models in Informative Chatbots

2023 | International Conference on Business Informatics | Citations: 0

Authors: Arz von Straussenburg, Arnold F.; Wolters, Anna

Abstract: Informative chatbots embedded in an organization-specific context can provide a ... Expand

Abstract: Informative chatbots embedded in an organization-specific context can provide a reliable, interactive, and engaging source of information for users. However, traditional chatbot techniques have limitations in processing and understanding user input and generating human-like responses. On the other hand, the latest implementations of large language models show promising results in these domains but have limitations in providing accurate and up-to-date facts from domain-specific knowledge bases. With the advent of popular chatbots like Chat-GPT, they are increasingly becoming part of organizations' digital infrastructure. In this research-in-progress paper, we argue that the strengths and weaknesses of traditional techniques and large language models are complementary. Therefore, we propose a hybrid chatbot architecture that utilizes inter-agent communication to compensate for disadvantages while enhancing the chatbot's abilities, as perceived by the user. This approach will form the basis for development and evaluation using Design Science Research (DSR) as part of our research. Collapse

Topics: information retrieval user experience

Methods: chatbot design artifact design science NLP method large language model

Towards a thematic dimensional framework of online fraud: An exploration of fraudulent email attack tactics and intentions

2023 | Decision Support Systems | Citations: 1

Authors: Bera, Debalina; Ogbanufe, Obi; Kim, Dan J.

Abstract: Despite anti-phishing filters, social engineering-based cyber-attacks still resu ... Expand

Abstract: Despite anti-phishing filters, social engineering-based cyber-attacks still result in billions of dollars lost annually, significant personal identity theft, loss of corporate secrets, and espionage. We review the phishing literature on psychological attacks and design tactics employed for deception by attackers. The result reveals the need to continuously identify tactics embedded in fraudulent email content to understand why users still fall prey to phishing attacks. Using the literature as a backdrop, we identify and conceptualize a thematic dimensional framework of fraudulent phishing attacks. This study uses benchmark datasets of fraudulent email to empirically validate the proposed dimensional framework. Specifically, a combination of machine learning-based content analysis and topic modeling found a majority of the discovered topics were labeled against the proposed di­ mensions. Statistical analysis was employed to provide empirical evidence of the thematic dimensions of fraudulent email attacks. This study thus, not only extends the cybersecurity literature by identifying and vali­ dating eight dimensions that enrich our understanding of phishing and attack identification but stimulates cu­ mulative research endeavors to develop yet more comprehensive dimensional frameworks of phishing email attacks. Collapse

Topics: phishing email phishing electronic mail email spam personalization

Methods: term frequency–inverse document frequency natural language processing topic model discriminant analysis literature study

A Theory-based Deep-Learning Approach to Detecting Disinformation in Financial Social Media

2023 | Information Systems Frontiers | Citations: 0

Authors: Chung, Wingyan; Zhang, Yinqiang; Pan, Jia

Abstract: The spreading of disinformation in social media threatens cybersecurity and unde ... Expand

Abstract: The spreading of disinformation in social media threatens cybersecurity and undermines market efficiency. Detecting disinformation is challenging due to large volumes of social media content and a rapidly changing environment. This research developed and validated a theory-based, novel deep-learning approach (called TRNN) to disinformation detection. Grounded in social and psychological theories, TRNN uses deep-learning and data-centric augmentation to enhance disinformation detection in financial social media. Temporal and contextual information is encoded as specific knowledge about human-validated disinformation, which was identified from our unique collection of 745,139 financial social media messages about four U.S. high-tech company stocks and their fine-grained trading data. TRNN uses multiple series of long short-term memory (LSTM) recurrent neurons to learn dynamic and hidden patterns to support disinformation detection. Our experimental findings show that TRNN significantly outperformed widely-used machine learning techniques in terms of precision, recall, F-score and accuracy, achieving consistently better classification performance in disinformation detection. A case study of Apple Inc.’s stock price movement demonstrates the potential usability of TRNN for secure knowledge management. The research contributes to developing novel approach and model, producing new information systems artifacts and dataset, and providing empirical findings of detecting online disinformation. Collapse

Topics: disinformation social media Twitter fake news usability

Methods: machine learning experiment design artifact deep learning recurrent neural network

Theories: psychological theory

Trust antecedents in online reviews across national cultures

2023 | Decision Support Systems | Citations: 0

Authors: Kim, Youngeui; Srite, Mark; Zhao, Huimin

Abstract: When reading an online review, consumers rely on implied trust-building factors ... Expand

Abstract: When reading an online review, consumers rely on implied trust-building factors within it. Trust antecedents have been extensively researched and categorized into argument quality and source credibility. However, there is still a need to investigate how their help for consumers differs across national cultures. Furthermore, while a profile photo has been found to enhance the helpfulness of a review by improving source credibility, little is known about how consumers perceive user appearance in a review picture and no research has compared these two types of visual information. To bridge these gaps, we examine how trust antecedents and visual information help consumers’ decision making differently across national cultures, using reviews collected from Amazon in the US and Germany, respectively. Our results show that information quality is more helpful for consumers in an individualistic and low uncertainty avoidance culture, whereas source credibility is more valued in the opposite culture. Interestingly, the effects of the two types of visual information, user appearance in a profile photo versus in a review picture, also differ across cultures. This research contributes to the online review literature by introducing visual information and national culture with new perspectives of trust. Our findings suggest that multinational e-commerce businesses adopt culture-specific strategies in review systems to better help consumers make informed purchase decisions. Collapse

Topics: national culture online review Vine data quality information retrieval

Methods: robustness check topic model term frequency–inverse document frequency propensity score method natural language processing

Introduction to the Minitrack on Intelligent Information Access and Retrieval

2023 | HICSS | Citations: 0

Authors: Pasi, Gabriella; Bordogna, Gloria

Abstract: The first edition of the minitrack on "Intelligent Information Access and Retri ... Expand

Abstract: The first edition of the minitrack on "Intelligent Information Access and Retrieval Systems" (IIARS) has been organized within the "Internet and the digital economy" track of HICSS 2023.We believe that the definition and development of IIARS is nowadays facing new challenges given both to the availability of big data on the Web and to the recent advances and application of machine learning and deep learning methods for Natural Language Processing and Information Retrieval. The minitrack has been conceived as a forum to present and discuss recent proposals and applications in the field of IIARS to trace novel emerging approaches and promising case studies. Collapse

Topics: user satisfaction information retrieval anonymity IT security

Methods: experiment machine learning natural language processing

Navigating digital transformation through an information quality strategy: Evidence from a military organisation

2023 | Information Systems Journal | Citations: 0

Authors: Struijk, Mylène; Angelopoulos, Spyros; Ou, Carol X. J.; Davison, Robert M.

Abstract: The use of digital technologies for extracting information from various data sou ... Expand

Abstract: The use of digital technologies for extracting information from various data sources can help organisations to reduce uncertainty and improve decision-making. The increasing availability in volume, velocity, and variety of data, however, can give rise to significant risks and challenges in ensuring a high level of information quality (IQ). Pre-digital organisations can be particularly susceptive to such challenges due to their limited experience with digital technologies and IQ governance. We adopt a theory-infused interventionist research approach to assist a pre-digital multinational military organisation in navigating its digital transformation (DT) by focusing on IQ. We design and implement an IQ strategy (IQS) by drawing upon organisational information processing theory and examining how the level of IQ can affect the balance between information processing requirements and capacity. We demonstrate that an IQS that incorporates both technological, as well as IQ governance solutions, can support organisations in setting the scope of their DT, decreasing employees' resistance to change, and increasing their satisfaction, while concurrently improving organisational efficiency. Our work stresses the importance of IQ in the digital era and delineates how pre-digital organisations can navigate DT by strategically addressing IQ. Collapse

Topics: data quality resistance to change cloud computing systems implementation data governance

Methods: qualitative interview personal interview action research archival research company material

Theories: organizational information processing theory information systems theory

Unifying Algorithmic and Theoretical Perspectives: Emotions in Online Reviews and Sales

2023 | Management Information Systems Quarterly | Citations: 2

Authors: Yu, Yifan; Yang, Yang; Huang, Jinghua; Tan, Yong

Abstract: Emotion artificial intelligence, the algorithm that recognizes and interprets v ... Expand

Abstract: Emotion artificial intelligence, the algorithm that recognizes and interprets various human emotions beyond valence (positive and negative polarity), is still in its infancy but has attracted attention from industry and academia. Based on discrete emotion theory and statistical language modeling, this work proposes an algorithm to enable automatic domain-adaptive emotion lexicon construction and multidimensional emotion detection in texts. Using a large-scale dataset of China's movie market from 2012 to 2018, we constructed and validated a domain-specific emotion lexicon and demonstrated the predictive power of eight discrete emotions (i.e., surprise, joy, anticipation, love, anxiety, sadness, anger, and disgust) in online reviews on box office sales. We found that representing overall emotions through discrete emotions yields higher prediction accuracy than valence or latent emotion variables generated by topic modeling. To understand the source of the predictive power from a theoretical perspective and to test the cross-culture generalizability of our prediction study, we further conducted an experiment in the U.S. movie market based on theories on emotion, judgment, and decision-making. We found that discrete emotions, mediated by perceived processing fluency, significantly affect the perceived review helpfulness, which further influences purchase intention. Our work shows the economic value of emotions in online reviews, generates insight into the mechanism of their effects, and has managerial implications for online review platform design, movie marketing, and cinema operations. Collapse

Topics: online review Python user-generated content microblog marketing management

Methods: computational algorithm word2vec parametric test product review experiment

Theories: psychological theory

A Knowledge Extraction Approach for IT Tech-support Transcripts

2023 | Americas Conference on Information Systems | Citations: 0

Authors: Zhao, Gary Yu; El-Gayar, Omar; Tu, Cindy Zhiling

Abstract: In information technology support/helpdesk transcripts, most of the data of int ... Expand

Abstract: In information technology support/helpdesk transcripts, most of the data of interest, such as the issue of concern, issue severity, context, and system status, is not provided in a structured form. Moreover, the special traits of product issue orientation, implicit background knowledge, and off-topic dialogues require a domain specialized approach to extract knowledge from these transcripts. Accordingly, this study analyzes the specific domain requirements and proposes a novel solution based on Natural Language Processing (NLP) approach. In the core process, this approach uses an adapted term frequency-inverse document frequency (TF-IDF) algorithm by adding a new parameter reflecting the term's priority in the text. Experimental results show that the proposed NLP-based solution performs reasonably well in topic categorizing with an accuracy of 92.8%. Compared to the performance of keywords extraction, the proposed approach achieves an accuracy of 93.4%, which outperforms the classic TF-IDF method signifying the importance of extracting and accommodating domain-specific knowledge. Collapse

Topics: system support Python open source application programming interface internet technology

Methods: natural language processing term frequency–inverse document frequency computational algorithm chatGPT experiment

Towards AI application marketplaces - an interview with Dorian Selz

2022 | Electronic Markets | Citations: 2

Authors: Alt, Rainer; Zimmermann, Hans-Dieter

Abstract: This interview with the CEO from Squirro reports on how a search engine in the e ... Expand

Abstract: This interview with the CEO from Squirro reports on how a search engine in the enterprise context evolves with artificial intelligence (AI) towards a  cognitive search engine and ultimately towards an electronic marketplace for information objects. Using several examples, Dorian Selz describes the opportunities of Squirro‘s low-code in creating user-specific information objects from internal and external data sources as well as in deriving probabilities and projections. This not only has the potential to improve existing business processes, but also sheds light on new predictive capabilities. The example shows potential elements of future AI marketplaces, such as models, app directories and connectors. Collapse

Topics: information retrieval low code development platform online learning system artificial intelligence electronic commerce

Let´s Do Our Bit: How Information Systems Research Can Contribute to Ethical Artificial Intelligence

2022 | Americas Conference on Information Systems | Citations: 0

Authors: Aslan, Aycan; Greve, Maike; Lembcke, Tim-Benjamin

Abstract: With the growing diffusion of Artificial Intelligence (AI), a variety of novel ... Expand

Abstract: With the growing diffusion of Artificial Intelligence (AI), a variety of novel ethical challenges become apparent. As a possible solution, the field of Ethical AI (EAI), the conviction, that to build trust in AI, ethical guidelines must be enforced, is gaining popularity. The importance of such guidelines has been recognized by regulatory institutions like the European Commission and led to the 'Artificial Intelligence Act', a legal framework to achieve EAI. In view of the growing importance of EAI and its implications for how societies interact with AI, it remains unclear to which extent information systems (IS) literature, socio-technical by its nature, can contribute to EAI. This article's contribution is threefold: It provides an explanation of the AI Act, introduces a systematic analysis of current IS literature on EAI, and derives future research directions for aspects of EAI which are currently insufficiently covered by IS scholars. Collapse

Topics: artificial intelligence privacy data governance database system information retrieval

Methods: literature study synthesis literature sample topic model qualitative content analysis