Search Results (4)

The rapid proliferation of Conversational Artificial Intelligence Assistants (CAIs) has transformed access to mental health information through freely accessible web interfaces, mobile applications, and public APIs (Application Programming Interfaces), yet systematic methodologies for their evaluation remain limited. This paper introduces SELCAI-MH, a multicriteria framework for CAI evaluation and selection. This framework integrates four complementary multicriteria methods: Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS), VIseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR), Complex Proportional Assessment Method (COPRAS), and Combinative Distance-based Assessment (CODAS), capturing distance-based, compromise-based, proportional, and negative-ideal logics, and proposes SOLAG, an aggregation method that produces a consensus ranking across methods. SELCAI-MH employs a dual evaluation mechanism combining psychiatric expert assessment with AI-based scoring, expert-derived criterion weights, and domain-relevant conversational datasets. The framework is applied to nine internet-accessible CAIs: proprietary platforms (ChatGPT 5.2, Claude Sonnet 4.5, Gemini 1.5 Flash, Perplexity Sonar, Bing AI/Copilot) and open-source Llama variants deployed via cloud inference endpoints. Using a set of anxiety-related questions and CAI responses, evaluated across seven criteria, Claude Sonnet 4.5 emerged optimal, followed by ChatGPT 5.2 and Gemini 1.5 Flash. SOLAG produced highly consistent rankings across the four multicriteria decision-making (MCDM) methods (Spearman ρ ≥ 0.98). Overall, SELCAI-MH provides a structured and reproducible decision-support framework for selecting accessible CAIs in sensitive mental health contexts. Full article

Artificial intelligence (AI) technologies, such as service robots, substantially influence frontline employees in the hospitality sector. This study highlights that conversational AI (CAI) may function as a viable outlet for hospitality workers to vent negative work-related issues. This function is particularly relevant because employees in this industry frequently experience customer mistreatment. Grounded in conservation of resources theory, we conceptualize venting to CAI as a resource-replenishing coping strategy triggered by customer mistreatment. Further, we theorize that this relationship is mediated by emotional exhaustion and moderated by trust in CAI, thereby strengthening the indirect effect. We collected and analyzed two-wave data from 394 frontline employees with CAI experience in the hospitality industry. The results indicate that customer mistreatment indirectly impacted frontline employees’ venting behaviors towards CAI, with emotional exhaustion functioning as the mediating mechanism. This indirect effect is particularly pronounced when employees exhibit high levels of trust in CAI. These findings offer practical insights for hospitality organizations aiming to leverage CAI as an accessible, low-risk tool for supporting employee emotional well-being and mitigating the negative consequences of customer mistreatment. Full article

As conversational artificial intelligence (CAI), including smart speakers, social robots, and dialogic learning apps, becomes increasingly present in children’s lives, its potential to support early mathematical thinking warrants closer attention. While existing research largely concentrates on literacy and language development, the role of CAI in early numeracy remains underexplored. This paper examines how voice-based CAI might contribute to informal mathematical thinking in early childhood. Adopting a conceptual lens, this paper synthesises existing theory and research to examine the potential roles of CAI in early mathematical learning. Grounded in sociocultural theory and dialogic pedagogy, this paper positions CAI as a potential mediator of early mathematical thinking through responsive dialogue. Four interrelated dimensions (child agency, cognitive scaffolding, mathematical language quality, and responsiveness and timing) are identified as a conceptual lens for evaluating how dialogue-based interactions with CAI may support or constrain young children’s mathematical thinking. Rather than framing CAI as a direct teaching tool, this paper explores its potential role as a dialogic partner in play-based numeracy and inquiry. The framework contributes to early mathematics education by highlighting both the promise and the limitations of CAI, offering guidance for research, technology design, and educational practice. It underscores the need for intentional, ethically informed integration of CAI that approximates the qualities of human dialogue while acknowledging current constraints in real-world use. Full article

Assessing the impact of land use and land cover (LULC) on water quality (WQ) is central to land-based environmental research. Adhering to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, this study analyses recent trends using LULC maps to assess stream WQ within buffers, focusing on papers published between 2020 and 2024. It identifies relevant remote sensing practices for LULC mapping, landscape metrics, WQ physicochemical parameters, statistical techniques for correlating LULC and WQ, and conventions for configuring buffers. Materials include Scopus, Web of Science, and Atlas.ti, which support both qualitative data analysis and Conversational Artificial Intelligence (CAI) tasks via its integration with OpenAI’s large language models. The methodology highlights creating a bibliographic database, coding, CAI, and validating prompts. Official maps and visual or digital interpretations of optical imagery provided inputs for LULC. Classifiers from earlier generations have shaped LULC cartography. The most employed WQ parameters were phosphorus, total nitrogen, and pH. The three most referenced landscape metrics were the Largest Patch Index, Patch Density, and Landscape Shape Index. The literature mainly relied on Redundancy Analysis, Principal Component Analysis, and alternative correlation approaches. Buffer configurations varied in size. CAI facilitated an agile systematic review; however, it encountered challenges related to a phenomenon known as hallucination, which hampers its optimal performance. Full article