The potential value of student-created mathematics learning objects

Learning and teaching.

A Design Framework for Personal Learning Environments

The development of learning objects (LO) has brought significant promise for creating highly personalized learning programs and it has resulted in the huge amount of various LOs all over the world. Learning object repository is regarded as a good median to manage the learning resources such as storing, indexing, retrieval and so on. However, due to the LOs' distribution and diversity, learners usually lack sufficient awareness and comprehension to make effective selection among the large numbers of available learning resources. It is difficult for the learners to properly integrate information to address their immediate learning need, for example, the related learning resources, or prerequisite learning domains, etc. In this paper, we put forward the idea of mining LO relation based on the learners' usage information in our LO repository (called as eLORM). By analyzing the LO relation patterns from their learning domain, granularity, categories, etc., the correlated various learning objects are able to be recommended to the learners. Based on the definitions of the LO relation patterns such as association pattern and sequence pattern, we apply the web usage mining methodologies in our LO relation patterns' model. We also develop a LO repository system based on our LO-relationship-mining approach.

Reusable learning objects (LOs) constitute a promising approach to the development of easily accessible, technologically sound, and curriculum aligned learning resources. Many research forums and scholarly articles have focused on the reusability of learning objects, metadata, and context issues, but few sources describe the economic challenges involved in implementing and sustaining an LO repository. What are the costs of establishing and maintaining a LO repository? Should funding for establishing and maintaining LO repositories come from institutional resources, consortium fees, grant money, LO sales, or other sources? To answer these questions we consider a variety of LO cost factors. We look at economic models used in distance education to see what they can tell us about LO economies. We discuss the relationship of funding approaches and operational scope (of a LO system) through considering a funding matrix that describes possible funding approaches. We discuss several emerging trends that may contribute to the future of learning resources from an economic perspective. Lastly, we provide several practical recommendations for funding LO repositories. In conclusion, we highlight developmental factors for LO repositories as they relate to the scope of operation and funding methods.

A learning object (LO) is a grouping of instructional materials structured to meet a specified educational objective. Digital LOs, which can be stored electronically, allow a new approach to instructional activity, making medical education more efficient, and potentially more cost-effective. They are reusable and can incorporate text, graphics, animations, audio, and video to support and enhance learning. A learning object can stand alone or be aggregated with additional objects to create larger forms of educational content meeting multiple educational objectives. Digital learning objects located in online repositories can be accessed by many computers and are easily handled by an array of learning management systems for delivery to learners at any time. Integrating digital learning objects with traditional educational methods in a blended learning approach assists medical educators in meeting the challenges of competing priorities. Multimedia LOs enable learners to tailor their experience to their preferred learning style. Through the use of learning objects, learners’ reactions, their acquisition of knowledge, skills and attitudes, and their behavioral changes become readily measurable. Learning objects provide multiple research opportunities, such as their use in adaptive learning, their added value in preclinical versus clinical education, and their impact as part of a blended learning strategy.Practice points•A learning object is a grouping of instructional materials structured to meet a specified educational objective.•Learning objects are interactive, interoperable, multimedia and easily accessible in online repositories.•Learning objects can stand alone or can be aggregated to create larger forms of educational content to meet multiple educational objectives.•Learning objects allow a blended learning approach to instructional activity, making medical education more efficient and potentially more cost-effective.•Learning objects create new teaching efficiencies for educators and create new learning opportunities.

Learning Objects (LOs) are web based learning resources presented by Learning Object Repositories (LOR). For recent years LOs have begun to take place on web and it is suggested that appropriate design of LOs can make positive impact on learning. In order to support learning, research studies recommends LOs should have been evaluated pedagogically and technologically, and the content design created by using LOs should have been designed through appropriate instructional models. Since the use of LOs have recently begun, an exact pedagogical model about efficient use of LOs has not been developed. In this study a LOR is designed in order to be used in mathematics education. The LOs in this LOR have been evaluated pedagogically and technologically by mathematics teachers and field experts. In order to evaluate the designed LO based environment, two different questionnaires have been used. These questionnaires are developed by using the related literature about web based learning environments evaluation criteria and also the items are discussed with the field experts for providing the validity. The reliability of the questionnaires is calculated cronbach alpha = 0.715 for the design properties evaluation survey and cronbach alpha =0.726 for pedagogic evaluation. Both of two questionnaires are five point Likert type. The first questionnaire has the items about “Learning Support of LOs, Competency of LOR, The importance of LOs in mathematics education, the usability of LOs by students”. “The activities on LOs are related to outcomes of subjects, there are activities for students have different learning styles. There are activities for wondering students.” are examples for items about learning support of LOs. “System helps for exploration of mathematical relations”, “I think teaching mathematics with this system will be enjoyable.” are example items for importance of LOs in mathematics education. In the competency of LOR title, “System can be used in problem design about daily life.”, “Using systems can take much time for teachers” and in the title Students‟ using of LOR, “System is not appropriate for collaboration”, “The learning environment can help communication between students.” are the example items. In the technological dimension, the title design principles consist of items like; “The text and tables on the LOs are readable. The design of the menus makes the system usable. , The design is original.” In technological consistency title “LOs can be found by search options.”, “The tools are used for users to continue on the system.” , “Many kinds of files can be uploaded in to the system.” and in security title, the questionnaire has items like; the upload and download systems do not have problems, the user control system works confident. By using these surveys 64 educators in the field and mathematics teacher evaluated the LOR in pedagogy and content dimensions, and 46 material design expert and web expert evaluated LOR in design, technology and security dimensions. As a result of pedagogical and content evaluation, the participants revealed positive views about the LO based learning environment. In pedagogical dimension; it is found that the number of LOs is enough, the content design system must be eliminated from the details, the LOs have enough interaction, LOs can help exploring mathematical relations, and the activities related affective domain must be increased and by using of LOs an exciting and funny learning environment may be designed. In addition; according to the design experts, the learning environment is basic and useful, also most of them agreed on the system that it is complied with design criteria. Besides experts evaluated the security options are feasible. By the results of this study, the LOR will be updated and revised to a form a web based learning environment for mathematics education and the real impact of LO based mathematics learning environment will be investigated in future studies.

<div class="page" title="Page 1"><div class="layoutArea"><div class="column"><p><span>The aim of this paper is to delineate a coherent framework for the authoring of re-purposable learning objects. The approach is orthogonal to the considerable work into learning object metadata and packaging conducted by bodies such as IMS, ADL and the IEEE. The 'learning objects' and standardisation work has been driven largely by adding packaging and metadata to pre-constructed learning artefacts. This work is very valuable. The argument of this paper, however, is that these developments must be supplemented by significant changes in the creation of learning objects. The principal aim of this paper is to delineate authoring principles for reuse and repurposing. The principles are based on a synthesis of ideas from pedagogy and software engineering. These principles are outlined and illustrated from a case study in the area of learning to program in Java.</span></p></div></div></div>

E-learning becomes a widely used form of educational service provided with or without teacher's support as a pre-study, main course, self-study support, or refreshment. A variety of course development tools, simple and free, or sophisticated and costly, are available for course development. Together with didactic-supported methodology and guidelines for learning content development, these instruments ensure a pedagogical quality of the resulting course. If a course development tool produces a learning content according to the SCORM standard [], it enables its usage as a whole or by some components - learning objects - by a large audience, i.e. those who can access it from a SCORM-compliant learning management system. Reusability, i.e. a capacity of being applied in various contexts, is an important feature for raising efficiency of learning content production and its ROI. So far, SCORM as a technical standard ensured technical side of reusability, whereas a didactically meaningful combination of the learning object was a responsibility of a course designer. Steady growth of users going on-line from their mobile devices - phones, smartphones, tablets, e-books - pushes mobile content development. Mobile learning research matured, growing from small sessions and workshops to an individual trend, resulting in a number of big international conferences. The research in m-learning emphasizes on its innovative and informal aspects, highlighting new learning activities and scenarios, enabling creativity, interaction and collaboration [ ]. Rough summary of m-learning could be visualized as isolated islands of innovative learning experiences, together with early pilots on mobile access to the web-based learning content and studies of basic content features accessible across various mobile platforms. Although some attention has been paid to the "legacy" content and its transfer to the mobile-accessible format, this issue has been addressed solely from a technical viewpoint. Thus, the quality of the resulting mobile learning was not a subject of study, and the differences in e-learning and m-learning environments were not addressed. This study is intended to extend the understanding of the reusability concept beyond technical solutions and emphasize on the value of didactically sound content in a similar way it is done by the Learning Design []. The ultimate goal may be seen in providing some guidelines for transformation of e-learning course to its m-learning version ensuring the preservation didactical quality. The purpose of this paper is to identify differences between e-learning and m-learning environments, and explore factors that may influence the quality of the e-learning content delivered in m-learning environment using a language course as an example. Khan's e-learning framework [ ] was selected as an instrument for the learning environment analysis. The proposed framework comprises eight dimensions, along which factors influencing the quality of the e-learning environment under study are analyzed. Since its introduction, the framework has been applied to various large-scale and small-scale environments. The language learning content is represented by the ELTEC - a self-study language course [ ] which is intended for enhancement of the professional tasks-oriented language skills. The course is highly interactive though some learning activities are preceded by the explanations and illustrations. Comparison of language learning activities within e-learning and m-learning environments led to some guidelines for creation of mobile version of the ELTEC. It was concluded that due to the environment differences and specifics of the mobile devices the course should be re-organized before being transformed to the mobile version, preserving the purpose but re-defining learning objectives.

In recent years, short video social platforms have garnered widespread attention, with Douyin (the Chinese version of TikTok) being a typical representative. The daily active users of the teen segment in Douyin are estimated to exceed 42 million. However, there still exists a certain blind spot in the research of Douyin's teen-specific content. This study conducts a content analysis of 360 short videos in Douyin's adolescent mode, and employs a Structural Equation Model (SEM) to analyze the relationships between the content quality, design features, learning objectives, age appropriateness, and user engagement. The results reveal that age appropriateness and learning objectives have a significant positive impact on content quality, and content quality and design features significantly positively impact user engagement. Further discussions point out that Douyin's adolescent mode faces challenges in dealing with issues like gender bias and algorithm bias, which may have adverse effects on the holistic development of teen users. The research findings provide empirical evidence for short video creators and platform optimization, and put forward targeted suggestions for the design of short video content in adolescent mode.

Students are given a set of prescribed learning objects within a domain area in a technology supported learning environment. Learning to be effective, the content should address the needs of students with various learning styles, promote adoption of a deep approach to learning, and help students to reach a higher level of intellectual gain. The current learning environment provides an opportunity for standardization of learning objects which can be reused, recombined and adapted through the use of LOs. The quality of the learning content delivered to a user also plays an important role and determination of such quality Learning Objects (LO) involves complex processes. An added knowledge to the existing LO can surely benefit the user in many ways and can improve the quality of an object. Metadata plays a bigger role in the delivery of Learning Objects based on user queries and many metadata standards exist which helps in retrieval of these LOs. A method is proposed to tag a new extended metadata called Knowledge Object (KO) of Knowledge Management System (KMS) with LOs. The LOs and KOs can automatically be classified using data mining approaches. K-nearest neighbour approach is used to find the closest KOs to a particular LOs These KOs are considered as extended metadata to fetch a Knowledge Enabled Learning Object as a specialized Knowledge Content to users; thereby the quality of an object in a learning environment is improved.

In today's world, WWW as a platform to deliver instruction enables e-learning systems to manage a large quantity of different digital content. Teachers contribute their expert knowledge and students contribute their investigation works, tasks, etc. As a consequence of this, the content repository grows with a lot of information. Then, an urgent need exists to retrieve the specific information they really need through e-learning systems. reusable learning object (LOs) concepts and standards for their application represent an advantage for knowledge management systems for e-learning systems. Users are able to manage and reuse content according to their needs without interoperability problems. However, the quality of the LOs content is not guaranteed. This paper attempts to analyze what must be considered to knowledge management for e-learning systems together with LOs and standars implication. A definition of LOs is proposed and quality criteria are suggested to evaluate LOs according to their characteristics and specific contexts.

The arrival of intelligent, general-purpose robots that can learn to perform new tasks autonomously has been promised for a long time now. Deep reinforcement learning, which combines reinforcement learning with deep neural network function approximation, has the potential to enable robots to learn to perform a wide range of new tasks while requiring very little prior knowledge or human help. This framework might therefore help to finally make general purpose robots a reality. However, the biggest successes of deep reinforcement learning have so far been in simulated game settings. To translate these successes to the real world, significant improvements are needed in the ability of these methods to learn quickly and safely. This thesis investigates what is needed to make this possible and makes contributions towards this goal. <br/><br/>Before deep reinforcement learning methods can be successfully applied in the robotics domain, an understanding is needed of how, when, and why deep learning and reinforcement learning work well together. This thesis therefore starts with a literature review, which is presented in Chapter 2. While the field is still in some regards in its infancy, it can already be noted that there are important components that are shared by successful algorithms. These components help to reconcile the differences between classical reinforcement learning methods and the training procedures used to successfully train deep neural networks. The main challenges in combining deep learning with reinforcement learning center around the interdependencies of the policy, the training data, and the training targets. Commonly used tools for managing the detrimental effects caused by these interdependencies include target networks, trust region updates, and experience replay buffers. Besides reviewing these components, a number of the more popular and historically relevant deep reinforcement learning methods are discussed.<br/><br/>Reinforcement learning involves learning through trial and error. However, robots (and their surroundings) are fragile, which makes these trials---and especially errors---very costly. Therefore, the amount of exploration that is performed will often need to be drastically reduced over time, especially once a reasonable behavior has already been found. We demonstrate how, using common experience replay techniques, this can quickly lead to forgetting previously learned successful behaviors. This problem is investigated in Chapter 3. Experiments are conducted to investigate what distribution of the experiences over the state-action space leads to desirable learning behavior and what distributions can cause problems. It is shown how actor-critic algorithms are especially sensitive to the lack of diversity in the action space that can result form reducing the amount of exploration over time. Further relations between the properties of the control problem at hand and the required data distributions are also shown. These include a larger need for diversity in the action space when control frequencies are high and a reduced importance of data diversity for problems where generalizing the control strategy across the state-space is more difficult.<br/><br/>While Chapter 3 investigates what data distributions are most beneficial, Chapter 4 instead proposes practical algorithms to {select} useful experiences from a stream of experiences. We do not assume to have any control over the stream of experiences, which makes it possible to learn from additional sources of experience like other robots, experiences obtained while learning different tasks, and experiences obtained using predefined controllers. We make two separate judgments on the utility of individual experiences. The first judgment is on the long term utility of experiences, which is used to determine which experiences to keep in memory once the experience buffer is full. The second judgment is on the instantaneous utility of the experience to the learning agent. This judgment is used to determine which experiences should be sampled from the buffer to be learned from. To estimate the short and long term utility of the experiences we propose proxies based on the age, surprise, and the exploration intensity associated with the experiences. It is shown how prior knowledge of the control problem at hand can be used to decide which proxies to use. We additionally show how the knowledge of the control problem can be used to estimate the optimal size of the experience buffer and whether or not to use importance sampling to compensate for the bias introduced by the selection procedure. Together, these choices can lead to a more stable learning procedure and better performing controllers. <br/><br/>In Chapter 5 we look at what to learn form the collected data. The high price of data in the robotics domain makes it crucial to extract as much knowledge as possible from each and every datum. Reinforcement learning, by default, does not do so. We therefore supplement reinforcement learning with explicit state representation learning objectives. These objectives are based on the assumption that the neural network controller that is to be learned can be seen as consisting of two consecutive parts. The first part (referred to as the state encoder) maps the observed sensor data to a compact and concise representation of the state of the robot and its environment. The second part determines which actions to take based on this state representation. As the representation of the state of the world is useful for more than just completing the task at hand, it can also be trained with more general (state representation learning) objectives than just the reinforcement learning objective associated with the current task. We show how including these additional training objectives allows for learning a much more general state representation, which in turn makes it possible to learn broadly applicable control strategies more quickly. We also introduce a training method that ensures that the added learning objectives further the goal of reinforcement learning, without destabilizing the learning process through their changes to the state encoder. <br/><br/>The final contribution of this thesis, presented in Chapter 6, focuses on the optimization procedure used to train the second part of the policy; the mapping from the state representation to the actions. While we show that the state encoder can be efficiently trained with standard gradient-based optimization techniques, perfecting this second mapping is more difficult. Obtaining high quality estimates of the gradients of the policy performance with respect to the parameters of this part of the neural network is usually not feasible. This means that while a reasonable policy can be obtained relatively quickly using gradient-based optimization approaches, this speed comes at the cost of the stability of the learning process as well as the final performance of the controller. Additionally, the unstable nature of this learning process brings with it an extreme sensitivity to the values of the hyper-parameters of the training method. This places an unfortunate emphasis on hyper-parameter tuning for getting deep reinforcement learning algorithms to work well. Gradient-free optimization algorithms can be more simple and stable, but tend to be much less sample efficient. We show how the desirable aspects of both methods can be combined by first training the entire network through gradient-based optimization and subsequently fine-tuning the final part of the network in a gradient-free manner. We demonstrate how this enables the policy to improve in a stable manner to a performance level not obtained by gradient-based optimization alone, using many fewer trials than methods using only gradient-free optimization.<br/>

Learning objects (LOs), generally understood as digital learning resources shared through the Internet and reused in multiple learning contexts, have aroused worldwide enthusiasm in the field of educational technology during the last years. Although LOs and LO systems offer tremendous possibilities to improve educational practices, there are many theoretical problems and practical shortcomings which are usually neglected. In this article we introduce the promises of cost‐effectiveness, reusability, modifiability and adaptability associated with LOs. Then we critically look at the problems underlying the LO approach, and provide our own alternative interpretation of LOs as useful elements of meaningful learning environments. Because of their flexible nature, LOs and LO systems can be used to support a variety of learning theories and instructional strategies—both sophisticated and reductionist ones (Parrish, 2004). LOs per se are not necessarily pedagogically functional or non‐functional, but their pedagogical value is determined through their context of use. The implementation of LOs requires a sound pedagogical grounding, and we argue that only by using LOs according to the principles of contemporary learning theories can their promises be fulfilled.

The development of e-learning cannot be separated from the creation of learning objects which are always improving and more diverse. In this study, we propose a technique to perform learning object searching that use a granular learning object concept using ontology representation. The concept of granular learning objects is done by separating the materials with their sub-materials. A semantic web-based search engine is also used to explore existing learning objects. The result of the research shows that the implementation of granular leaning object ontology, and semantic web-based search engine, can improve searching performance by 29,7p better than search engine which uses document index-based searching technique. The search results do not only show the general description of a learning object, but can also show the sub-materials and the relationships among those materials. An information in the form of pictures and their descriptions about the retrieved learning object can also be displayed.

BackgroundThe World Health Organization (WHO) Patient Safety Curriculum Guide defines learning objectives for patient safety. Current implementation in healthcare education is insufficient. Possible explanations may be obsolescence and/or a shift in needs. We investigated whether overarching topics and specific learning objectives of the WHO Patient Safety Curriculum Guide are still up-to-date, their attributed importance, and their perceived difficulty to achieve.MethodsExperts on patient safety and medical education from 3 European countries were asked to suggest learning objectives concerning patient safety using group concept mapping. Following 3 successive steps, experts rated ideas by importance and difficulty to achieve. Correlation analyses investigated the relationship between those. Overarching topics of the learning goals (clusters) were identified with multivariate analysis.ResultsA total of 119 statements about intended learning objectives on patient safety were generated, of which 86 remained for sorting and rating. Based on multivariate analyses, 10 overarching topics (clusters) emerged. Both the learning objectives and the overarching topics showed high correspondence with the WHO Patient Safety Curriculum Guide. Strong correlations emerged between importance and difficulty ratings for learning objectives and overarching topics.ConclusionsThe WHO Patient Safety Curriculum Guide’s learning goals are still relevant and up-to-date. Remarkably, learning objectives categorized as highly important are also perceived as difficult to achieve. In summary, the insufficient implementation in medical curricula cannot be attributed to the content of the learning goals. The future focus should be on how the WHO learning goals can be implemented in existing curricular courses.