AI-Driven Digital Learning Recognition: Evaluating User Perceptions of Beedik Implementation

This study explores user perceptions of a digital system for AI-based learning recognition called Beedik. As artificial intelligence (AI) transformations in the education system increasingly take hold, understanding how learners interact with the recognition platform becomes crucial to ensuring the system's usability, trustworthiness, and pedagogical value. Using a quantitative survey method, data were collected from 134 users and analyzed through descriptive statistics and PLS SEM. Using a descriptive quantitative approach, perceptions were measured across the dimensions of Access, Clarity, Reliability, Relevance, and Satisfaction. Measurement model testing demonstrated the instrument's reliability and validity (all Cronbach's alpha, rho_A, and Composite Reliability ≥ 0.70; AVE ≥ 0.50), confirming that the indicators consistently reflect the constructs. In the structural model, Reliability and Access had the strongest and most significant influence on satisfaction, but Clarity and Relevance showed a positive but not yet significant direction. These findings emphasize the development priority of system reliability and ease of access, followed by improving clarity and relevance. Implicatively, the study contributes to the literature on educational technology acceptance and supports the lifelong education agenda through systematically designed digital transformation.本研究针对一款名为Beedik的人工智能（AI）驱动学习识别数字系统的用户感知展开探索。随着人工智能（AI）在教育体系中的变革愈发深入，明晰学习者与该识别平台的交互逻辑，对保障系统的可用性、可信度与教学价值至关重要。本研究采用定量调查法，从134名用户处采集数据，并通过描述性统计与偏最小二乘结构方程模型（PLS SEM）开展分析。本研究采用描述性定量研究框架，从可及性（Access）、清晰性（Clarity）、可靠性（Reliability）、相关性（Relevance）与满意度（Satisfaction）五个维度对用户感知进行测量。测量模型检验结果证实，研究工具具备优异的信度与效度：所有克朗巴哈α系数、rho_A值与组合信度均≥0.70，平均变异萃取量（AVE）≥0.50，表明各指标可稳定映射对应的构念。在结构模型中，可靠性与可及性对满意度的影响最为显著且强度最高，而清晰性与相关性则呈现正向关联但未达到统计显著水平。本研究结果明确了系统可靠性与可及性的优先开发方向，其次需优化平台的清晰性与相关性。从理论贡献层面而言，本研究丰富了教育技术接受领域的相关文献，并通过系统化设计的数字化转型，为终身教育议程提供了实践支撑。