Modern online learning: problems and trends
Today, online learning is beginning to really push out the full-time form of learning. Even some progressive Russian universities are slowly introducing online courses designed to replace couples in schools. But how much is all this real? Today we talk about the problems and trends of modern online learning.
And which is more convenient for you: go to class or study at home, wrapped in a blanket?
')
I give the word to the author.
In this article we would like to share with you thoughts about the problems and trends of modern online learning, as well as the possibility of creating VR projects that can be integrated with modern e-learning.
In the modern world, the concept of e-Learning is complex and includes several components:
If we consider the evolution of online learning in the context of sociocultural theories, then development proceeded from behaviorism to social constructivism. Initially, the training system was based on behavioral principles - the teacher was located in the center, and students around him silently listening to him.
The transition to more complex learning models began with the work of Lev Vygotsky, who suggested that learning occurs as a result of social interaction with other people. Constructivist theories arose from this thesis. They assume that knowledge is not external to the student, namely, each student builds his own model of some specific knowledge. When using these models, the focus shifts from teacher to student, which draws knowledge from a variety of sources.
A few more learning concepts are connective and social models. They assume that knowledge is a network of various interrelated sources. Accordingly, the amount of knowledge increases as sources are added to the network. Sources can be, for example, people or books, and the Internet.
The question arises: “How to implement all forms of education?” Traditionally, if we look at the development of e-Learning, then huge systems and protocols were created under the behavioral concept (instruction based). But the point was that all systems were built under the presence of a certain package of knowledge. It can be formed in the form in which it is necessary for integration into a particular system.
Historically, the SCORM protocol was initially used extensively in the US Air Force and it was the behaviorist training system that was intended. This implied that the student’s actions were completely unimportant - he should simply memorize all the information provided. However, SCORM becomes obsolete, and over time it becomes less and less effective. Now the question is more relevant: “How is it possible in the modern world to organize a student-centric system of education using exactly online services?”
To answer this question, it is necessary to consider three theses:
To take into account the first thesis, you need to use individual learning paths. However, with this approach it becomes necessary to obtain a huge amount of data on the student's behavior.
To solve the problem specified in the second thesis, we come to the idea about the features of the current SCORM and AICC protocols. They assume work in online, that is, if the training takes place in the classroom or classroom, then their use is automatically impossible.
The main problem in the field of social education, as indicated by the third thesis, is the lack of development. However, now they talk a lot about this type of training, but no one really knows how to implement it correctly. The time has come to introduce new standards for learning protocols. Replaced by SCORM xAPI.
The main idea behind the development of xAPI is the separation of statistics from the content of the learning material itself. This is necessary for more flexible options for collecting these statistics themselves. For example, if the learning material does not exist in electronic form, let's say the student receives information using a book.
Another paradigm is simulation, it always stood alone. Simulation is the environment in which a student can do whatever he wants. However, in order to achieve the desired result, the person is given some input, but he chooses the “route” of the simulation. Currently, all VR-trainings and serious games work outside of training systems almost without data exchange. xAPI allows you to integrate data exchange in VR training, which is not possible using the SCORM protocol.
Traditionally, e-learning systems have evolved as web-based systems. Accordingly, the development vector went from Web 1.0 to Web 2.0 - that is, to create custom content. And this paradigm is very suitable for learning systems, primarily due to the fact that it gave teachers quite simple means to create educational content.
If you look at any Learning Management System now, then it represents quite good means for creating presentations, various tests and everything you need to develop a training course, all this being done directly in the browser. Of course, there are quite a few third-party authoring tools, and they are actively used, especially for creating some more complex content - but at the end of the chain we still have the Internet and publication in a format suitable for embedding in a web page.
But the development of various systems for working with 3D content was completely different. First of all, not all games (and the games were long enough, and remain now, one of the engines of progress in this area) required a network. And especially - the network is complicated. Moreover, the means for creating custom content is not something new (level editors are already very many years old), but the means for creating a fundamentally new functionality, deep modding is already noticeably younger. It is traditionally considered that the modder knows what he is doing, that is, he is at least a little bit, but he understands programming, has an idea of ​​the pipeline to work with 3D models and other specific things.
But in this case this is not very suitable, since it is required to create a training system, and make it accessible to all. That is, it is necessary to simplify and at the same time complicate “modding,” making it more similar to the Web 2.0 model, without losing its multiuse. Let's try first to look at how a standard MMO game usually looks, and think about what is missing in such an architecture.
If you look at (from a very large height, without details) on how any multiplayer 3D space is arranged, for example, standard MMO since WoW, then we will see something like the following:
Naturally, there is a client and server. The main task of the server is to synchronize the actions of various users. To do this, a certain fixed protocol of communication between the client and the server is defined, through which we can send messages from the client to the server and, in the simplest case, send to all other clients. (Of course, there is a huge layer of nuances associated with the implementation of this process - for example, is the server authoritarian, how is the lag compensated, how is work with physics, etc. - but at this level of the scheme you should skip it). In addition, you need to somehow track the logic of interactions on the server side, for example, game scripts. To do this, you will most likely need some server-side scripting. To simplify life, scripting on the client side is also very useful.
There is, however, the question: “Where does the content itself, that is, models, textures, sounds, and everything else, come from in this scheme?”. All this is “inside” the client. That is why MMO clients are often so heavy in terms of size. When something new is added, users need to download the update, and only the developer, his programmers and artists can add something new.
The question arises: “What needs to be changed in the architecture to enable users to create content?”
Immediately there is an idea to store content on the server side. In this embodiment, the client looks more like a web browser - it downloads everything you need and caches.
In this direction several projects have been made, the most famous of which, perhaps, is Second Life. The use of such a scheme in its pure form is not always justified - sometimes a hybrid approach is better when part of the content is downloaded directly (for example, when a new scene is loaded), and part is loaded dynamically.
The next question is what to do with logic. All the same Second Life offered a fairly unique server-side scripting scheme at that time — the user wrote the code for the behavior of the object in the embedded client editor, and then the script was sent to the server and executed on it. But there are a number of questions here too. First, the script must have access to some key server subsystems, and the question arises - how many such subsystems should there be and what should they be like? Secondly, there is the question of what language it should be and how its standard library should look like - LSL, the Second Life scripting language, was rather primitive and complicated at the same time (for example, there was no normal work with arrays and there was no OOP in principle).
The concept is somewhat similar to SL, but differs in a number of details. An element of a virtual environment can be a full-fledged application that is compiled and has all the necessary logic of working “inside” itself, i.e. a kind of black box from the point of view of the entire client-server system and other applications. At the same time, the entire virtual environment can be the possible, but not the only, interface of the application. It also complements the standard library, providing a number of functions for working with yourself. In other words, this is an analogue of an ordinary application in a modern operating system, just the OS function is performed by the environment itself.
In addition to the interactions themselves in 3D space, the virtual application has access to the client interface and the ability to create their own interface parts. Otherwise, ideally, the application should not be aware of the division into client and server. The code should be written in the same way as a regular single-user application would be written.
The virtual environment, the objects of which are such applications, we call dynamic virtual environment.
There are, however, a number of quite obvious questions about the practical implementation of such a system - how to combine these principles with reliability, performance and many other criteria, and how to implement a specific architecture of such a project? We will tell about it in the following article.
Jedium is a Microsoft partner company working in the field of virtual, augmented reality and artificial intelligence. Jedium has developed a framework to simplify the development of complex projects on Unity, some of which are publicly available on GitHub . Jedium plans to replenish the repository with new modules of the framework, as well as integration solutions with Microsoft Azure.
Vitaly Chaschin - Software developer with more than 10 years of experience in the design and implementation of three-dimensional client-server applications - from concept to full implementation and integration of applications and virtual reality solutions. Systems Architect Jedium LLC, MSc in IT.
Alexey Sarafanov
marketing manager at Jedium LLC.
Sergey Kudryavtsev
CEO and founder of Jedium LLC.
And which is more convenient for you: go to class or study at home, wrapped in a blanket?
')
I give the word to the author.
In this article we would like to share with you thoughts about the problems and trends of modern online learning, as well as the possibility of creating VR projects that can be integrated with modern e-learning.
In the modern world, the concept of e-Learning is complex and includes several components:
- Blended learning is a combination of training with an instructor and online training with non-class options. A student can create projects, use the help of mentors and so on. Blended learning is synchronous and asynchronous (the synchronous version implies an immediate feedback on the progress of the teacher. Asynchronous uses the concept of home independent tasks);
- Mobile learning - Learning using mobile devices;
- Informal learning (informal learning) - activities outside the formal environment (class, online class, etc.). This type of training works through social interaction.
If we consider the evolution of online learning in the context of sociocultural theories, then development proceeded from behaviorism to social constructivism. Initially, the training system was based on behavioral principles - the teacher was located in the center, and students around him silently listening to him.
The transition to more complex learning models began with the work of Lev Vygotsky, who suggested that learning occurs as a result of social interaction with other people. Constructivist theories arose from this thesis. They assume that knowledge is not external to the student, namely, each student builds his own model of some specific knowledge. When using these models, the focus shifts from teacher to student, which draws knowledge from a variety of sources.
A few more learning concepts are connective and social models. They assume that knowledge is a network of various interrelated sources. Accordingly, the amount of knowledge increases as sources are added to the network. Sources can be, for example, people or books, and the Internet.
The implementation of concepts
The question arises: “How to implement all forms of education?” Traditionally, if we look at the development of e-Learning, then huge systems and protocols were created under the behavioral concept (instruction based). But the point was that all systems were built under the presence of a certain package of knowledge. It can be formed in the form in which it is necessary for integration into a particular system.
Building a student-centered learning system and a new learning protocol
Historically, the SCORM protocol was initially used extensively in the US Air Force and it was the behaviorist training system that was intended. This implied that the student’s actions were completely unimportant - he should simply memorize all the information provided. However, SCORM becomes obsolete, and over time it becomes less and less effective. Now the question is more relevant: “How is it possible in the modern world to organize a student-centric system of education using exactly online services?”
To answer this question, it is necessary to consider three theses:
- The system should be student centric;
- The system should support blended learning;
- Training in the system should be with social experience.
To take into account the first thesis, you need to use individual learning paths. However, with this approach it becomes necessary to obtain a huge amount of data on the student's behavior.
To solve the problem specified in the second thesis, we come to the idea about the features of the current SCORM and AICC protocols. They assume work in online, that is, if the training takes place in the classroom or classroom, then their use is automatically impossible.
The main problem in the field of social education, as indicated by the third thesis, is the lack of development. However, now they talk a lot about this type of training, but no one really knows how to implement it correctly. The time has come to introduce new standards for learning protocols. Replaced by SCORM xAPI.
xAPI
The main idea behind the development of xAPI is the separation of statistics from the content of the learning material itself. This is necessary for more flexible options for collecting these statistics themselves. For example, if the learning material does not exist in electronic form, let's say the student receives information using a book.
Another paradigm is simulation, it always stood alone. Simulation is the environment in which a student can do whatever he wants. However, in order to achieve the desired result, the person is given some input, but he chooses the “route” of the simulation. Currently, all VR-trainings and serious games work outside of training systems almost without data exchange. xAPI allows you to integrate data exchange in VR training, which is not possible using the SCORM protocol.
Construction of virtual spaces
Traditionally, e-learning systems have evolved as web-based systems. Accordingly, the development vector went from Web 1.0 to Web 2.0 - that is, to create custom content. And this paradigm is very suitable for learning systems, primarily due to the fact that it gave teachers quite simple means to create educational content.
If you look at any Learning Management System now, then it represents quite good means for creating presentations, various tests and everything you need to develop a training course, all this being done directly in the browser. Of course, there are quite a few third-party authoring tools, and they are actively used, especially for creating some more complex content - but at the end of the chain we still have the Internet and publication in a format suitable for embedding in a web page.
But the development of various systems for working with 3D content was completely different. First of all, not all games (and the games were long enough, and remain now, one of the engines of progress in this area) required a network. And especially - the network is complicated. Moreover, the means for creating custom content is not something new (level editors are already very many years old), but the means for creating a fundamentally new functionality, deep modding is already noticeably younger. It is traditionally considered that the modder knows what he is doing, that is, he is at least a little bit, but he understands programming, has an idea of ​​the pipeline to work with 3D models and other specific things.
But in this case this is not very suitable, since it is required to create a training system, and make it accessible to all. That is, it is necessary to simplify and at the same time complicate “modding,” making it more similar to the Web 2.0 model, without losing its multiuse. Let's try first to look at how a standard MMO game usually looks, and think about what is missing in such an architecture.
Different MMO generations and architecture approaches
If you look at (from a very large height, without details) on how any multiplayer 3D space is arranged, for example, standard MMO since WoW, then we will see something like the following:
Naturally, there is a client and server. The main task of the server is to synchronize the actions of various users. To do this, a certain fixed protocol of communication between the client and the server is defined, through which we can send messages from the client to the server and, in the simplest case, send to all other clients. (Of course, there is a huge layer of nuances associated with the implementation of this process - for example, is the server authoritarian, how is the lag compensated, how is work with physics, etc. - but at this level of the scheme you should skip it). In addition, you need to somehow track the logic of interactions on the server side, for example, game scripts. To do this, you will most likely need some server-side scripting. To simplify life, scripting on the client side is also very useful.
There is, however, the question: “Where does the content itself, that is, models, textures, sounds, and everything else, come from in this scheme?”. All this is “inside” the client. That is why MMO clients are often so heavy in terms of size. When something new is added, users need to download the update, and only the developer, his programmers and artists can add something new.
The question arises: “What needs to be changed in the architecture to enable users to create content?”
Immediately there is an idea to store content on the server side. In this embodiment, the client looks more like a web browser - it downloads everything you need and caches.
In this direction several projects have been made, the most famous of which, perhaps, is Second Life. The use of such a scheme in its pure form is not always justified - sometimes a hybrid approach is better when part of the content is downloaded directly (for example, when a new scene is loaded), and part is loaded dynamically.
The next question is what to do with logic. All the same Second Life offered a fairly unique server-side scripting scheme at that time — the user wrote the code for the behavior of the object in the embedded client editor, and then the script was sent to the server and executed on it. But there are a number of questions here too. First, the script must have access to some key server subsystems, and the question arises - how many such subsystems should there be and what should they be like? Secondly, there is the question of what language it should be and how its standard library should look like - LSL, the Second Life scripting language, was rather primitive and complicated at the same time (for example, there was no normal work with arrays and there was no OOP in principle).
DVO
The concept is somewhat similar to SL, but differs in a number of details. An element of a virtual environment can be a full-fledged application that is compiled and has all the necessary logic of working “inside” itself, i.e. a kind of black box from the point of view of the entire client-server system and other applications. At the same time, the entire virtual environment can be the possible, but not the only, interface of the application. It also complements the standard library, providing a number of functions for working with yourself. In other words, this is an analogue of an ordinary application in a modern operating system, just the OS function is performed by the environment itself.
In addition to the interactions themselves in 3D space, the virtual application has access to the client interface and the ability to create their own interface parts. Otherwise, ideally, the application should not be aware of the division into client and server. The code should be written in the same way as a regular single-user application would be written.
The virtual environment, the objects of which are such applications, we call dynamic virtual environment.
There are, however, a number of quite obvious questions about the practical implementation of such a system - how to combine these principles with reliability, performance and many other criteria, and how to implement a specific architecture of such a project? We will tell about it in the following article.
The authors
Jedium is a Microsoft partner company working in the field of virtual, augmented reality and artificial intelligence. Jedium has developed a framework to simplify the development of complex projects on Unity, some of which are publicly available on GitHub . Jedium plans to replenish the repository with new modules of the framework, as well as integration solutions with Microsoft Azure.
Vitaly Chaschin - Software developer with more than 10 years of experience in the design and implementation of three-dimensional client-server applications - from concept to full implementation and integration of applications and virtual reality solutions. Systems Architect Jedium LLC, MSc in IT. Alexey Sarafanovmarketing manager at Jedium LLC.
Sergey KudryavtsevCEO and founder of Jedium LLC.
Source: https://habr.com/ru/post/418111/
All Articles