随着互联网技术的迅速发展，以及新冠疫情期间全球范围内超大规模的在线教育实践 的开展，在线学习已经成为教育体系常态化不可或缺的一部分。在线学习给人们带来更丰 富的学习资源、更自由灵活的学习方式，允许随时随地开展学习 ， 更能发挥学习者的认知 主体性。然而，在实际的教学实践中，在线学习仍面临一些重要挑战，比如学习者自主调 节学习能力弱、学习投入低、学习浅层化严重，协作学习的效果不理想、学习者参与协作 贡献的积极性不高等。而有研究指出社会性交互是在线学习成功的关键要素，有效的社会 性交互可以促进概念转变，并提升学习效果。为此 ，近年来越来越多的研究者开始关注如 何设计适应性服务，如同伴服务、适应性协作分组服务等，以提升在线学习的社会性交互 水平。然而已有研究中这种聚焦单一时刻、服务要素之间彼此孤立的适应性服务，难以满 足在学习过程中学习者与知识内容、学习同伴等进行持续性、多维度交互的学习需求，而 具有面向学习过程、动态演化生成、连接人与知识等特征的社会知识网络能够为上述学习 需求的满足提供新的可能。

在社会学习相关理论和支持策略的指导下，本研究基于社会知识网络，面向在线学习 中的个体学习过程和群体协作学习过程，设计和实现了两种适应性服务形态——社会化导 航服务和学习圈子服务，并探索了其应用效果。随后，在实证研究结论的基础上，进一步 面向在线多场景社会性学习，构建了基于社会知识网络的适应性服务模型，以支持学习者 之间有意义的社会性交互，促进自我调节学习和知识建构，从而提升在线学习效果。具体 说来，主要开展了以下三个方面的研究 ：

第一，基于社会知识网络的适应性服务形态设计与实现 。基于社会知识网络，设计和 实现了具有关联性、适应性、可视化、支架式、多态性特征的在线学习适应性服务形态， 包含面向个体学习过程的社会化导航服务和面向群体协作学习过程的学习圈子服务。其中， 社会化导航服务，以社会认知视角 下 的自我调节学习理论为依据，并整合社会调节学习工 具设计原则和社会比较策略，形成 了关联学习过程中的人和内容资源的可视化网络服务， 实现了以“策略性社会比较“为原则的、有效促进自我调节学习的适应性支架式服务 ； 学 习圈子服务，以跨越界限的学习发生机制为理论依据，并整合了群体感知策略，形成了关 联人、内容和生成性人工制品的动态网络，实现了以“求同存异”和“水平多样”为原则 的、促进集体知识建构的适应性、可视化的支架式服务。

第二，基于社会知识网络的适应性服务应用效果实证研究 。为了检验适应性服务的应 用效果，本研究选取某高校本科生专业课的两个自然教学班，分别开展了两个教学实验。 前者探究了社会化导航服务对自我调节学习的影响，后者探究了学习圈子服务对集体知识建构的影响。实验一的结果表明：社会导航服务能够有效提升自我调节学习的成效、显著 提升学习过程中的自我监测水平，对于其社会比较取向为表现取向的学习者，还能有效提 升其使用深度学习的策略、显著改善自我调节的整体感受；在学习行为方面，社会化导航 服务能够帮助学习者更加积极地与他人建立连接，有效减轻其对教师的依赖，更积极地参 与学习活动和自我调节。 实验二的结果表明：学习圈子服务能够在领域知识构建的整体范 围、分化程度和知识的具象化水平方面，有效提升集体知识建构的成效，促进积极的社会 情感过程，尤其是在培养团队凝聚力方面作用显著；在学习行为方面，学习圈子服务能够 显著提升群体学习者之间的交互紧密性、整体社会网络向中心聚集的程度，并能够有效拓 展学习者之间社会性连接的范围，增强学习者对学习过程的感知和创造性参与，有助于支 持更高社会性水平的交流、协作和创新实践。

第三，基于社会知识网络的适应性服务模型构建。从跨界学习的理论视角，基于实证 研究的结论，构建了基于社会知识网络的适应性服务模型。模型的核心要素包含一个社会 知识网络、两种适应性服务形态、四类服务支架、五种跨界学习场景。其中，一个社会知 识网络是对在线学习交互过程的动态语义化表征；两种适应性服务形态，指的是基于社会 知识网络实现的社会化导航服务和学习圈子服务，具有关联性、适应性、支架式、可视化、 多态性等特征，遵循 “策略比较＂、“求同存异”和“水平多样”的适应性原则，包含学习 同伴、学习小组、知识导航、认知特征、参与特征、社会身份特征多种相互关联的服务要 素。这两种适应性服务形态，可以面向五种跨界学习场景，提供四类服务支架。五种跨界 学习场景分别为个体跨界学习、组内成员之间的跨界学习、小组与小组间的跨界学习、小 组与社群间的跨界学习、社群之间的跨界学习。依据这五种跨界学习的学习发生机制和对 服务支架的需求，模型描述了两种适应性服务形态可以提供的四类服务支架，即支持自我 调节、促进知识建构、拓展社会连接、提供服务定制。

基于上述研究工作的开展，本研究在两个方面取得了 一定的创新。

(1) 在理论创新方面：本研究面向学习过程，提出了具有关联性、适应性、支架式、 可视化、多态性特征的两种适应性服务形态一一社会化导航服务和学习圈子服务——及其 设计原则和要素，进而构建了基于社会知识网络的适应性服务模型，可为五种跨界学习场 景提供四类服务支架，弥补了已有研究在服务形态孤立、服务支架缺乏等方面的不足。

(2) 在实践创新方面，本研究依托学习元乎台设计和实现了基于社会知识网络的适应 性服务支撑系统，并基于此在真实教学场景下开展了两个教学实验，证实了适应性服务对 自我调节学习和集体知识建构的正向影响效果，形成了两个创新教学案例，为适应性服务 的设计、实现及应用提供了实践方案。

With the rapid development of the Internet technology and the large-scale online education practice that took place worldwide during the COVID-19 pandemic, online learning has become an indispensable part of the normalization of education system. Online learning brings people more abundant learning resources, freer and more flexible learning methods, allowing learning to be carried out anytime and anywhere, and bringing into play to learners' cognitive subjectivity. However, in actual teaching practice, online learning still encounters some major difficulties, such as learners' inadequate ability to self-regulated learning, low learning engagement, serious shallowness of learning, unsatisfactory effect of collaborative learning, and low motivation of participating in collaborative contribution. Studies have pointed out that effective social interaction can promote conceptual change and improve learning outcomes, and it has been identified as a critical element of successful online learning. Therefore, in recent years, more and more researchers have begun to focus on how to design adaptive services, such as peer services and adaptive collaborative grouping services, so as to enhance social interaction in online learning. However, the adaptive services in existing studies, which focus on single moments and isolate service elements from each other, can hardly meet the learning needs of continuous and multidimensional interaction between learners, knowledge content and learning peers in the learning process. Social knowledge network with the characteristics of learning process oriented, dynamic evolution generation, and connection between human and knowledge can provide new possibilities for meeting the above learning needs.

Under the guidance of social learning theories and supporting strategies, this study designs and implements two forms of social knowledge network-based adaptive service, including social navigation service and learning circle service, for individual learning process and group collaborative learning process in online learning. And it explores their application effects through empirical research. Subsequently, on the basis of the empirical research conclusions, an adaptive service model based on social knowledge network is constructed for online multi-scene social learning, in order to support meaningful social interaction between learners, promote self-regulated learning and knowledge construction, so as to improve the effect of online learning. Specifically, research has been carried out in the following three aspects.

First, the design and implementation of adaptive service forms based on social knowledge network. Based on the social knowledge network, the adaptive service forms of online learning with the characteristics of relevance, adaptability, visualization, scaffolding and polymorphism were designed and implemented, including the social navigation service for individual learning process and the learning circle service for group collaborative learning process. Specifically, the social navigation service is based on the self-regulating learning theory from the perspective of social cognition, and integrates the design principles of social regulating learning tools and social comparison strategies to form a visual network service of learner and content resources in the process of related learning. It also realizes the adaptive scaffolding service based on the principle of strategic social comparison to effectively promote self-regulating learning. The learning circle service, based on the boundary crossing learning mechanism, integrates the group perception strategy, forms a dynamic network of people, content and generative artifacts, and realizes the adaptive and visual scaffolding service that promotes the construction of collective knowledge based on the principle of "seeking common ground while shelving differences" and "diverse levels".

Second, empirical research on the application effect of social knowledge network-based adaptive service. In order to validate the application effect of the designed adaptive service, this study selected two natural teaching classes of specialized courses of undergraduate students in a university and carried out two experiments respectively. The former experiment explores the influence of social navigation services on self-regulated learning, while the latter explores the influence of learning circle services on the construction of collective knowledge. The results of the first experiment show that social navigation services can effectively improve the effectiveness of self-regulated learning, significantly improve the level of self-monitoring in the learning process, and effectively improve the use of deep learning strategies and the overall feeling of self-regulation for the learners whose social comparison orientation is performance orientation. In terms of learning behavior, social navigation services can help learners to establish connections with others more actively, effectively reduce their dependence on teachers, and more actively participate in learning activities and self-regulation. The results of the second experiment show that learning circle service can effectively improve the effectiveness of collective knowledge construction and promote the positive social emotional process, especially in the cultivation of team cohesion in terms of the overall scope of domain knowledge construction, the degree of differentiation and the level of knowledge visualization. In terms of learning behavior, learning circle service can significantly improve the interaction closeness among group learners, the degree of the overall social network to the center, and effectively expand the range of social connections between learners, enhance the perception of learners and creative participation in the learning process, and help support the communication, collaboration and innovative practice at a higher social level.

Third, an adaptive service model was constructed based on social knowledge network. From the theoretical perspective of boundary crossing learning and based on the conclusion of empirical research, the adaptive service model based on social knowledge network was constructed. The core elements of the model include one social knowledge network, two adaptive service forms, four types of service scaffolds, and five types of cross-border learning scenarios. Specifically, the social knowledge network is a dynamic semantic representation of the interactive process of online learning. The two adaptive service forms refer to the social navigation service and the learning circle service based on the social knowledge network. They are characterized by relevance, adaptability, scaffolding, visualization and polymorphism, and follow the adaptive principles of "strategy comparison", "seeking common ground while reserving differences" and "level diversity". It includes multiple interrelated service elements such as learning partner, learning group, knowledge navigation, cognitive characteristics, participation characteristics and social identity characteristics. These two adaptive service forms can provide four types of service scaffolds for five kinds of cross-border learning scenarios. The five types of cross-boundary learning scenarios are individual cross-boundary learning, inter-group cross-boundary learning, inter-group cross-boundary learning, inter-group cross-boundary learning, inter-group cross boundary learning, and inter-community cross-boundary learning. According to the learning mechanism of these five types of cross-border learning and the demand for service scaffolds, the model describes four types of service scaffolds that the two adaptive service forms can provide, namely supporting self-regulation, promoting knowledge construction, expanding social connection, and providing service customization.

Based on the above research work, this research has achieved certain innovations in two aspects.

First, theoretical innovation: in this study, two adaptive service forms, namely social navigation service and learning circle service, with the characteristics of relevance, adaptability, scaffolding, visualization and polymorphism, were proposed for the learning process and their design principles and elements, and then an adaptive service model based on social knowledge network was constructed, which can provide four types of service scaffolds for five kinds of boundary crossing learning scenarios. It makes up for the shortcomings of the existing researches in the isolation of service forms and the lack of service supports.

Second, practical innovation: this study designed and implemented an adaptive service support system based on social knowledge network in Learning Cell platform. Two experiments were then carried out based on this adaptive service support system in real teaching scenarios, which confirmed the positive effects of adaptive service on self-regulated learning and collective knowledge construction, and formed two innovative teaching cases. It also provides a practical scheme for the design, implementation and application of adaptive service.