Multi-Turn, Open-Domain, Task-Oriented, and Knowledge-Based Dialogue Systems for Adaptive Workplace Learning and Organizational Knowledge Support: A Systematic Review and Meta-Analysis

This project presents a systematic review and meta-analysis of empirical research on multi-turn, open-domain, task-oriented, and knowledge-based dialogue systems developed to support adaptive workplace learning and organizational knowledge support. The review aims to synthesize evidence on dialogue-system architectures, workplace learning applications, datasets, evaluation methodologies, and reported effectiveness while identifying methodological limitations and future research opportunities. Modern organizations increasingly require employees to learn continuously while performing work tasks. Dialogue systems have emerged as promising technologies for providing conversational access to organizational knowledge, supporting professional training, facilitating knowledge retrieval, and delivering context-aware performance support. Despite rapid advances in conversational artificial intelligence, the literature remains fragmented across multiple disciplines, including artificial intelligence, natural language processing, human-computer interaction, information systems, and knowledge management. Existing studies differ considerably in their dialogue architectures, datasets, evaluation metrics, and experimental protocols, making cross-study comparison difficult and limiting evidence-based guidance for researchers and practitioners. This review addresses these challenges through a structured evidence synthesis consisting of a systematic review and a quantitative meta-analysis where sufficient comparable evidence exists. Objectives The project seeks to: 1. Identify and classify dialogue-system approaches that support adaptive workplace learning and organisational knowledge support. 2. Examine the datasets, evaluation metrics, and assessment methodologies used across empirical studies. 3. Quantitatively evaluate the effectiveness of dialogue-system approaches where comparable evidence is available. 4. Identify methodological limitations, recurring challenges, and research gaps to guide future research and practice. Research Questions RQ1. What types of multi-turn, open-domain, task-oriented, and knowledge-based dialogue systems have been investigated to support adaptive workplace learning and organisational knowledge support? RQ2. Which datasets, evaluation metrics, and assessment approaches are most commonly used to evaluate dialogue systems in workplace learning, organisational learning, knowledge management, and performance-support contexts? RQ3. How effective are different dialogue-system approaches in supporting workplace learning, knowledge retrieval, performance support, and personalised learning interactions? RQ4. What limitations, methodological challenges, and research gaps exist in the evaluation and deployment of dialogue systems for adaptive workplace learning and organisational knowledge support?