Supporting community-based learning: case study of a geographical community organization designing its website

Community computing supports human - computer interaction among neighbours in geographical or place-based community organizations. Using a case study of such an organization, we investigate the process of designing their website. Our long-term participatory design approach, integrating developmental informal learning, allowed us to understand how this community organization adopts, evaluates, and sustains website technology. Based on our case study analysis, we present three design heuristics for developing community-based technology: align and afford new possibilities for participation, dynamically manage organizational knowledge and learning, and enhance social capital within community organizations and with the broader community.     

In search of online pedagogical models: investigating a paradigm change in teaching through the School for All community

Following a preliminary evaluation of the Web‐mediated School for All in 2002, this study further examines the online teacher role and explores the possible pedagogical models in this Web‐based informal lifelong learning context through a long‐term innovative e‐learning project over a 2‐year period. The author documents the educational rationale, reviews related literature on Web‐based instructional design, and presents longitudinal in‐depth study results based on empirical data collected from the four Web‐course contests held over 2 years from the perspective of outstanding online teachers. Significantly, three categories of six Web‐based pedagogical models induced from this research project are presented, and the rationale for how and why teachers would prefer to use certain teaching models is discussed. In addition, the characteristics of the outstanding online teachers and issues inherent in this alternative mode of a Web‐based informal educational setting are discussed. Finally, more research issues, questions, and possibilities for future research work are raised or identified for further study.     

Churchman's Hegelian Inquiring System and Perspectival Thinking

This paper focuses upon a critical aspect implicit in C West Churchman's Hegelian Inquiring System, namely its ability to provide all the necessary ingredients for the emergence of a learning culture. This focus is achieved by articulating certain phenomenological aspects of the Hegelian model. In particular the relationship between subject and object is explored. After its emergence the learning culture then requires a special kind of thinking, which results in certain capacities, to continue its evolvement. This kind of thinking has been defined as Perspectival Thinking and is elaborated in the context of an Information System as a social system bound to technology. In the absence of a learning culture no organization can sustain its learning capacities, yet how often are learning cultures taken for granted? This paper explores a number of ideas that provide the ground for developing and sustaining learning cultures.     

基于深度学习的专利知识推荐服务研究

专利作为一种包含大多数领域中最完整的设计信息,可以为设计者解决设计问题提供有价值的指导。针对现有的专利推荐方法难以有效地推荐跨领域专利的问题,提出一种基于深度学习的跨领域专利知识推荐方法,用于创新产品的概念设计。对产品功能和知识需求情境进行建模,将设计问题进行标准化表达,生成设计问题空间。提出一种半监督学习算法（TG-TCI）将专利功能信息按照功能基自动分类和标记,利用实体识别算法（BERT-BiLSTM-CRF）提取专利应用场景术语、技术术语,结合国际专利分类（IPC）信息以表示专利的功能、情境、技术和领域属性,从而生成专利知识空间。通过设计问题空间到专利知识空间的功能基和知识情境映射查找所需的跨领域专利,根据技术和领域属性对它们进行聚类和评估,选出特定的专利以激发设计者的创造力。以一个实际案例进行分析验证,证明了基于深度学习的专利知识推荐模型的可行性及有效性。     

Contextual perspective in analysing collaborative knowledge construction of two small groups in web-based discussion

This paper presents a methodology designed to explore the role of context in collaborative knowledge construction activity in asynchronous web-based discussion. The discussions of two student groups participating in a web-based teacher education course were compared. The comparison aimed to highlight the differences and similarities between the groups’ knowledge construction activity through studying the thematic structure, communicative functions and contextual resources used in their discussions. The results indicated that the different backgrounds of the two student groups influenced the way context was created and interpreted, and how meanings were negotiated. The differences and similarities between the groups’ activity illuminated the situated and mediated nature of learning. The possibilities of the methodology used in this study for evaluating collaborative knowledge construction in context are also discussed.     

深度学习算法在问句意图分类中的应用研究

在聊天机器人多轮对话中如何根据上下文理解用户的意图是多轮交互中的一个重点问题，也是一个难点问题。现有的问句理解方法大多是针对单句的，且侧重于某种句式结构的理解。如何根据上下文语境对当前用户的意图进行识别，而不仅仅是针对单轮进行一个个分析，使得对话在一个连续的语境下具备细粒度的理解能力，是一个亟待解决的问题。针对以上问题，提出了一种基于深度学习的自然语言问句多意图分类方法，其中涉及到的用户意图包含闲聊类、音乐类、新闻类、算术类、餐饮类、订票类、天气类、服务类等13类。首先使用自然语言处理的相关技术对多轮对话进行处理分析，识别出其中的关键词，然后使用深度学习方法和分层分类技术构建了二分类和多分类深度学习模型，学习上下文语境和语义关系，共同对用户意图进行识别。通过实验证明了构建的深度学习模型对用户意图识别的准确率分别为94.81%、93.49%。因此，所提方法基本能够解决自然语言问句意图识别的问题。     

基于深度学习的遥感图像茶园区域识别应用研究

得益于遥感技术的发展和深度学习在图像处理方面的进展,采用深度学习识别遥感图像的方法被广泛应用。与传统的统计农作物种植面积方法相比较,通过深度学习的方法来识别茶园种植区域,可以减少人工依赖,节约人力资源,实时获取数据,具有更高的时效性。数据来源于Bigmap,以贵州省卫星遥感图像为数据基础,提出了使用深度学习来识别茶园区域的应用方法。实验目标为从整张遥感图像中提取出茶园种植区域。首先对遥感图像进行数据预处理,然后采用人工目视解译的方法标注出茶园区域并制成数据集,将数据集导入神经网络进行训练获得网络模型,最后将验证图像放入到训练好的神经网络当中,获得验证结果;检测精确率为95.83%,检测召回率为85.00%。     

Moving towards engaged learning in STEM domains; there is no simple answer,but clearly a road ahead

What is the best approach to educating students is, evidently, the pivotal question in educational research. In the general debate on this question, clear positions are often taken―for example, whether teacher‐led instruction or more student‐directed approaches should be followed. In the current narrative review, this central question and the different stances taken are explored for a specific field of learning, STEM (science, technology, engineering, and mathematics) domains. The current article starts with an argument as to why the traditional, more directive teacher‐led educational approach to STEM learning may not always lead to deep conceptual knowledge and proposes exploring more engaged forms of learning. More specifically, a particular arrangement for engaged learning, inquiry learning, is taken as an example in this exploration. It is argued that the effects of educational methods are influenced by a multitude of (interacting) factors related to student and teacher characteristics, context, domain, frequency of use, timing, and probably more. Therefore, this article tries to outline a more nuanced and balanced stance towards choosing educational approaches. A picture is presented to show that adopting and combining different approaches, at both the lesson and the curricular levels, may be necessary to reach optimal levels of learning. Existing and upcoming technologies may play a decisive role in realizing this more complex, but also more realistic view of learning in STEM.     

深度域适应综述:一般情况与复杂情况

信息时代产生的大量数据使机器学习技术成功地应用于许多领域.大多数机器学习技术需要满足训练集与测试集独立同分布的假设,但在实际应用中这个假设很难满足.域适应是一种在训练集和测试集不满足独立同分布条件下的机器学习技术.一般情况下的域适应只适用于源域目标域特征空间与标签空间都相同的情况,然而实际上这个条件很难满足.为了增强域...     

Crossing the digital divide safely and trustingly: how ecologies of learning scaffold the journey

The article addresses the issue of learning to elearn in borderless programs in a globalised learning landscape and the associated problems of scaffolding the journey across the digital divide. The authors argue that the assumption underlying such courses is that cross-cultural programs are viable because they are conceived and designed to be ‘global’, and that they assume this design to be inclusive. Henning and Van der Westhuizen claim that the global discourse in most domains can take only marginal note of the need to infuse such programs with a local semiotic—a course design criterion for which they argue. They furthermore forward the notion that the majority of the world's prospective elearners need various bridging mechanisms in order to be able to access the broader discourse and that one of these mechanisms can be explored through the metaphor of “information ecologies” as proposed by Nardi and O'Day [Nardi, B.A., & O'Day, V.L. (1999). Information ecologies. Using technology with heart. Cambridge, MA: The MIT Press]. They also conclude that issues of the learners’ trust in the course and its system need to be considered when contemplating programs for diverse target groups. By way of a case study, consisting of three portraitures of adult learners, they explore the limitations of assumed distributed cognition and claim that learning is, in reality, contained/constrained in the familiar local narrative of the novice adult elearners in a rural South African context. The case study illustrates how the resistance to technology and its power base becomes an obstacle for the students and how the support of peers becomes the main scaffolding mechanism for their entry into electronic learning environments. The findings thus show how the social context becomes the facilitator and the scaffold for elearning, more than technology and the curriculum itself.     

深度学习在智能机器人中的应用研究综述

机器人发展的趋势是人工智能化,深度学习是智能机器人的前沿技术,也是机器学习领域的新课题。深度学习技术被广泛运用于农业、工业、军事、航空等领域,与机器人的有机结合能设计出具有高工作效率、高实时性、高精确度的智能机器人。为了增强智能机器人在各方面的能力,使其更智能化,介绍了深度学习与机器人有关的研究项目与深度学习在机器人中的各种应用,包括室内和室外的场景识别、机器人的工业服务和家庭服务以及多机器人协作等。最后,对深度学习在智能机器人中应用的未来发展、可能面临的机遇和挑战等进行了讨论。     

After the Paradigm of Contemporary Physics in Architecture: Spatial Possibilities and Variations

Architecture in the information age becomes a mixture of spatial and temporal processes that are directly linked to changes in science, technology and culture. In the digital era, when information becomes a matter of philosophy, methods of informational transitions provide multiple possibilities for conceptualizing space, and thus opening new horizons to architecture. This research explores algorithms of analogy between string theory, M-theory and architecture. The mathematical spacetime concepts of movement, dimension and topology in physics are studied as tools for achieving informational transitions in the design elements and their structures. As a result this methodology opens numerous possibilities for interpretation and creation of new design morphologies.     

基于深度学习的情境感知推荐系统研究进展

传统的推荐系统存在数据高度稀疏、冷启动及用户偏好建模难等问题,而把情境信息融入推荐系统中能有效缓解此类问题.深度学习技术已经成为人工智能领域研究热点,把深度学习应用在情境感知推荐系统当中,为推荐领域的研究带来新的机遇与挑战.本文从情境感知推荐系统相关概念出发,综合整理国内外研究相关文献,介绍深度学习技术融入情境感知推荐系统相关应用模型,提出了基于深度学习的情境感知推荐系统研究的不足以及对未来的展望.     

深度学习在图像识别中的应用研究综述

深度学习作为图像识别领域重要的技术手段，有着广阔的应用前景，开展图像识别技术研究对推动计算机视觉及人工智能的发展具有重要的理论价值和现实意义，文中对深度学习在图像识别中的应用给予综述。介绍了深度学习的由来，具体分析了深度信念网络、卷积神经网络、循环神经网络、生成式对抗网络以及胶囊网络等深度学习模型，对各个深度学习模型的改进型模型逐一对比分析。总结近年来深度学习在人脸识别、医学图像识别、遥感图像分类等图像识别应用领域取得的研究成果并探讨了已有研究值得商榷之处，对深度学习在图像识别领域中的发展趋势进行探讨，指出有效使用迁移学习技术识别小样本数据，使用非监督与半监督学习对图像进行识别，如何对视频图像进行有效识别以及强化模型的理论性等是该领域研究的进一步方向。     

Learning object design considerations for small-screen handheld devices

The key limitation of handheld technology for the delivery of learning objects is the small screen that is available for effective display. The smallness of the screen not only adversely affects the clarity, but it also negatively impacts on the acceptance and integration of this potentially useful technology in education. Handheld devices are likely to change further in size in the future with consumer demand for less bulky but more powerful devices. This exploratory study investigated characteristics of effective design of learning objects on such devices. This paper reports upon user response to learning object design possibilities and provides a set of recommendations to guide improved utility and future research.