67471 research outputs found
Sort by
Pack and Measure: An Effective Approach for Influence Propagation in Social Networks
Get PDFInfluence Maximization has been widely used to enhance the effectiveness of online marketing campaigns. In this paper, we consider the Influence Maximization problem under the Independent Cascade model (IC). The problem asks for a minimum set of nodes in a network to serve as seed set from which a maximum influence propagation is expected. New seed-set selection methods are introduced based on the notions of a d-packing and node centrality. In particular, we focus on selecting seed-nodes that are far apart and whose estimated influence values are the highest in their local communities. Our best results are achieved via an initial computation of a d-Packing followed by selecting either nodes of high degree or high centrality in their respective closed neighborhoods. This overall “Pack and Measure” approach proves highly effective as a seed selection method. Our results have direct practical implications on conducting an online marketing campaign
Understanding Human Companionship with Artificial Intelligence: Insights from Replika-related Information Systems Research
Get PDFThe emergence of social chatbots designed to simulate emotionally supportive relationships constitutes a substantial advancement in human-technology interaction. Among these, Replika has emerged as the most salient and contentious example, garnering considerable and sustained scholarly attention within the Information Systems (IS) community. Scholars have investigated the processes by which individuals establish and cultivate companionship with Replika, as well as the broader implications of such interactions. Nevertheless, this corpus of knowledge remains fragmented, impeding a comprehensive understanding of what user’ interactions with Replika elucidate about human-AI companionship. This paper undertakes a systematic review of IS literature that centres specifically on Replika, with the objectives of consolidating extant insights and proposing avenues for future research
The Volatility of Microwork: Daily Working Conditions, Worker Well-Being, and the Role of Motivation
Get PDFMicrowork is a growing market of the gig economy, contributing to the development of technology through human data processing services. This work is highly unpredictable and is conducted under precarious conditions, which impacts workers’ well-being. We expected daily fluctuations in microworking conditions to promote worker well-being contingent on workers’ motivation. Using a daily diary design, we examined how three microwork-specific conditions affect workers’ general well-being. Our findings reveal that daily fluctuations in working conditions directly impact workers’ psychological well-being. Further, these working conditions affect psychological well-being, psychological distress, and psychosomatic well-being through workers’ basic need satisfaction. However, we did not find that workers’ motivation moderated the association between working conditions and basic need satisfaction. This study illustrates that the volatility of working conditions in microwork plays a distinct role in shaping workers’ well-being. We call for changes at task- and organization-level of platforms to create more sustainable working environments
Initial Review of Goal-Directed Answer Set Programming for Orchestration of Neuro-Symbolic UAV Autonomy
Get PDFAutonomous unmanned aerial vehicles (UAVs) are capable of solving a variety of problems in environments where humans cannot easily go, such as military zones or search and rescue operations. Edge computing explores the trade-off between running software at the edge onboard a UAV (limited space for processing power and memory) versus on a server (increased latency and connectivity problems). The state-of-the-art solution for UAV path planning is the use of deep learning, however these systems are large and unexplainable. Furthermore, when these models need to be adjusted they often cannot be retrained on the edge. We propose the use of a powerful symbolic reasoner that can be run and modified on the UAV for path planning, making explainable and trustworthy edge autonomous systems. We evaluate the feasibility of this approach with a proof-of-concept symbolic UAV system based on the VECSR system for commonsense reasoning, which we call VECSR-
Probabilistic Soft Logic for Toxic Intent Prediction in Conversation: a Moral Foundations Theory-Driven Neural-Symbolic Framework
Get PDFTransformer-based models (RoBERTa, DeBERTa, ToxicBERT) achieve high accuracy but suffer critical limitations: they operate as black boxes, classify content based on surface features regardless of toxic intent, and exhibit biases when profanity is present. We present a neural-symbolic framework that integrates Probabilistic Soft Logic with Moral Foundations Theory, enabling accurate prediction of toxic intent and moral reasoning. Our approach improves prediction accuracy, provides interpretable explanations, and demonstrates synergistic effects. We evaluated 748,283 toxic instances (from a total of 1,268,154 conversations) across Telegram war discourse and Reddit climate debates. The results show a 4.9 percentage point improvement over the strongest transformer baseline with statistical significance. The framework provides interpretable explanations for 91.3\% of predictions through PSL rules grounded in moral psychology. Ablation analysis confirms genuine neural-symbolic synergy with full integration, achieving F1=0.847, significantly outperforming individual neural and symbolic components. This advances ethical AI and conversation moderation
Empowering International Students’ Listening and Speaking Skills via AI-Driven Audio in NotebookLM: Reflections Across Disciplines
Get PDFDespite the growing interest in GenAI, there has been limited focus on its multimodal potential, moving beyond text-based applications to auditory learning. One promising avenue is using AI-generated podcasts based on self-selected academic literature, enabling learners to engage with content in new ways. Research suggests that students, especially those who are not native English speakers, often struggle with academic speaking skills. This paper explores using NotebookLM. Learners can listen to a podcast generated and practice discussions with the AI-hosts to deepen their understanding of key concepts and prepare for live seminars. Drawing on pedagogical reflections from teachers of international students, this paper explores how this technology can support diverse learners in engaging with academic literature and developing academic speaking skills, as well as the potential concerns associated with its use. The paper concludes with recommendations for educators on better supporting international students through multimodal, AI-enhanced learning tools
AI in International Marketing: How Digital Service Firms Standardize and Adapt Marketing Across Borders
Get PDFThis study explores how digital service firms apply Artificial Intelligence (AI) in their international marketing efforts. Specifically, we examine how these firms use mechanical, thinking, and feeling AI to standardize, adapt, and personalize their international marketing to international customers. Based on a qualitative multiple-case study, including five Finnish digital service firms, we find that mechanical AI helps different marketing processes related to standardization of marketing mix elements. Thinking AI, commonly used together with mechanical AI, is applied to facilitate the adaptation of product, promotion, and pricing strategies for foreign markets and customers. Feeling AI is applied to personalize the marketing offer based on individual customer needs. Based on these findings, we suggest three propositions that advance research on this topic, and which indicate how each type of AI relates to different stages of the marketing process and various marketing mix elements. This study contributes to research and practice in international marketing by offering a detailed view of how digital service firms apply AI in in their global marketing efforts
Meaning Matters for Large Language Models
Get PDFLarge language models (LLMs) have achieved remarkable adoption. While AI providers position their systems as assistive helpers and knowledge tools, users increasingly employ them for open-ended interactions seeking life-advice or creative exploration. This raises questions about the role of LLMs in such meaning-making activities, and the extend to which LLMs can access and encode meaning. In this conceptual essay, we apply Paul Ricoeur's hermeneutic philosophy to distinguish between structural and existential forms of meaning, revealing that LLMs can function as sophisticated conversational partners capable of engaging their vast “text” while lacking access to experientially-grounded understanding. We come to interpret user prompting as genuine hermeneutic encounters that enable meaning-making. We further reveal that hallucinations, the propensity of LLMs to generate plausible sounding yet incorrect responses, represent inevitable architectural trade-offs rather than eliminable technical failures. Our framework suggests new directions for LLM design that embrace generative capabilities and establishes principles for responsible user engagement
Adaptive Machine Learning for Dynamic Environments: Evaluating Data Drift-Triggered Retraining in COVID-19 Severity Prediction
Get PDFMachine learning (ML) models deployed in operational systems often face performance degradation due to data drift. Designing effective, scalable strategies for adaptive model maintenance remains an open challenge. This study evaluates four retraining strategies—naïve (static), periodic, clinically guided context-driven, and data drift-triggered—in a dynamic, high-stakes environment: predicting severe COVID-19 outcomes. Using a large-scale CDC dataset, we find that data drift-triggered retraining strikes an effective tradeoff between predictive performance and retraining cost. It matches the performance of periodic retraining while requiring far fewer retraining cycles and offers a fully automated mechanism for adaptive learning. In contrast, context-driven retraining performs well and requires fewer retraining cycles but depends on expert input and lacks automation. Our findings provide empirical insights and practical guidance for designing adaptive ML systems in dynamic environments, with implications for both researchers and practitioners deploying ML models in healthcare and other domains with evolving data landscapes