随着以互联网为代表的信息技术的飞速发展，人们的学习方式发生了巨大的变化，人人可学、处处可学、时时可学成为现实，社会联通、去中心化、开放共享、动态变化的学习方式逐渐形成。这种转变使学习者不再是局限于被动接受知识的传统群体，而是通过扮演多种角色并灵活转换，成为自主学习者。与此同时，以 cMOOC 为代表的联通主义学习形式应运而生并受到广泛关注。在联通主义学习中，学习者同样较大的学习自主权，他们能够根据自身学习水平和能力、选择合适的学习环境，通过灵活实施学习行为，实现角色扮演与转换。同时，联通主义学习倡导通过群体智慧的汇聚解决模糊、复杂、没有确定答案的问题，这对于以问题解决为目的的学习具有重要意义。然而当前却鲜有研究关注到联通主义学习中的问题解决过程，更缺乏对这一学习过程中学习者角色及其转换的认识。

基于此，本研究的核心问题是：在联通主义学习学习情境中，从问题解决的视角来看，学习者群体扮演的角色及其转换的特征和规律有哪些，具体包括三个问题：联通主义学习情境中：（1）学习者在问题解决中扮演了哪些角色？不同角色有何特征？（2）学习者群体扮演的角色在问题解决中是如何转换的？其转换在不同问题解决过程和不同学习阶段中有何差异？（3）具有不同身份的促进者扮演了哪些学习者角色？同时在角色转换中发挥了什么作用？

基于以上研究问题，本研究首先对相关概念进行了界定，包括：联通主义学习情境、问题与问题解决、学习者角色。其次，基于文献调研中确定的联通主义教学交互模型和问题解决过程模型并采用理论演绎的方法构建了适用于本研究的学习者角色模型，该模型包括规则迷茫者、信息探索者等9类学习者角色及其行为解释和操作性指标。

进而本研究选取用国内第一门cMOOC《互联网+教育：理论与实践的对话》7.0作为研究对象，基于本研究所建立的学习者角色模型，运用内容分析法和滞后序列分析法对10周学习期间学习者（包含促进者和非促进者）产生的2074条交互文本进行了分析，重点从同一问题解决过程和跨问题解决过程两个角度对学习者在整体学习期间的角色类型与转换进行了分析，同时分析了促进者在其中的作用；其次分析了不同阶段学习者角色的类型与转换的特征以及促进者在其中的作用。经过分析，最终得出以下5点结论：（1）联通主义学习情境中，讨论协商者、话题启发者、信息连接者是进行问题解决时出现最频繁的学习者角色；（2）学习者角色转换遵循非线性的问题解决过程，同一学习者角色存在于多种转换路径之中；（3）在不同学习阶段，学习者角色转换路径既有共性，又呈现出差异性；（4）相较而言，话题启发者和信息连接者是促进者在问题解决过程中扮演的最突出的角色；（5）随着学习阶段的深入，促进者所扮演的学习者角色类型逐渐减少，但在学习者角色转换中的作用持续扩大。

本研究主要有两个创新点：一是构建了适用于分析联通主义学习情境中基于问题解决的学习者角色编码框架，这将使得后续研究者可以更深入地探究联通主义学习中的问题解决过程；二是基于实证数据发现了联通主义学习情境中基于问题解决的学习者角色类型及其转换特征，这将有助于联通主义课程设计者更好地设计和优化课程，从而为学习者提供个性化的教育支持，以促进学习者在问题解决过程中的角色转换，从而提高学习效果。

With the rapid development of information technology, represented by the internet, people's learning methods have undergone significant changes. Ubiquitous learning, characterized by accessibility, decentralization, open sharing, and dynamic changes, has gradually taken shape. This transformation has enabled learners to transition from traditional passive recipients of knowledge to autonomous learners, who can play various roles and flexibly switch between them. Concurrently, connectivism learning, represented by cMOOCs, has emerged and attracted widespread attention. In connectivism learning, learners have substantial autonomy, enabling them to choose appropriate learning environments and engage in flexible learning behaviors, thus facilitating role-playing and transitions. Connectivism learning also advocates for collective intelligence to address ambiguous, complex, and indefinite problems, which is of vital significance for problem-solving-oriented learning. However, current research scarcely focuses on the problem-solving process in connectivism learning, let alone the understanding of learners' roles and their transitions.

In light of this, the core question of this study is: In connectivism learning contexts, from the perspective of problem-solving, what are the characteristics and patterns of learner role transitions? This inquiry includes three specific questions: In connectivism learning contexts, (1) What roles do learners play in problem-solving, and what are the features of these roles? (2) How do learner roles transform during problem-solving, and what are the differences in these transitions across various problem-solving processes and learning stages? (3) What learner roles do facilitators with different identities play, and what is their role in role transitions?

Based on these research questions, this study first defines relevant concepts, including connectivism learning contexts, problems and problem-solving, and learner roles. Next, the study constructs a learner role model applicable to this research, based on the connectivism teaching interaction model and problem-solving process model identified through literature review. This model comprises nine learner roles, such as rule-seekers and information explorers, along with their behavioral explanations and operational indicators.

Subsequently, this study selects the first domestic cMOOC course, "Internet + Education: Dialogue between Theory and Practice" 7.0, as the research object. Using content analysis and lag sequence analysis, the study analyzes 2,074 interaction texts generated by learners (including facilitators) during the 10-week learning period, focusing on the role types and transitions of learners in the overall learning process from the perspectives of both single and cross-problem solving processes. The study also analyzes the role of facilitators in different learning stages. The analysis yields the following five conclusions: (1) In connectivism learning contexts, discussion negotiators, topic initiators, and information connectors are the most frequent learner roles during problem-solving; (2) Learner role transitions follow a nonlinear problem-solving process, with a single learner role appearing in multiple transition paths; (3) In different learning stages, learner role transition paths share commonalities while also exhibiting differences; (4) Compared to other roles, topic initiators and information connectors are the most prominent roles played by facilitators in the problem-solving process; (5) As the learning stage progresses, the types of learner roles played by facilitators tend to decrease, but their role in learner role transitions continuously expands.

This study has two main innovations: First, it constructs a learner role coding framework applicable to the analysis of connectivism learning contexts based on problem-solving, enabling subsequent researchers to delve deeper into the problem-solving process in connectivism learning. Second, it discovers the types and transition characteristics of learner roles in connectivism learning contexts based on problem-solving using empirical data, which can help connectivism course designers better design and optimize courses, thereby providing personalized educational support for learners. This, in turn, promotes learner role transitions in the problem-solving process and ultimately improves learning outcomes.