The paper considers some new methods of using blockchain technology as a tool of improving educational process management. Two e-learning problems are described. The rst one is low ratings of completing MOOCs. The second one is the need for more detailed and accurate reports. Blockchain technology and its applications in education is considered. Blockchain's ability of documenting events is reviewed particularly. Two cases of using blockchain in this way are performed. The rst one is making educational reports more detailed and easier to design. Another one is using blockchain for creating any student model in adaptive education to make electronic educational courses more e ective and to increase ratings of completing MOOCs. A blockchain's option of documenting timestamps and users' addresses automatically is specially mentioned. Both cases are described mathematically and united in one model. Using Ethereum blockchain for the evaluation because of its smart-contract's mechanism is justi ed. Sequence diagram showing application work ow, is presented. Input data from the real e-learning course are processed with using statistics to get the most frequently used system events. Application tools for working with the Ethereum blockchain are presented. The process of developing, deploying and mining smart-contract is shown. Output report is presented. Implementation is described, the ways of future learn are reported.
Today online education becomes more and more massive [1{3]. There are a lot
of students, therefore, a lot of data. To analyze and control educational process,
we need a lot of parameters to handle. Every learning management system has
a module for generating and analyzing reports. But as MOOCs become more
popular, new methods of manage and control educational process are required.
1 Copyright c 2019 for this paper by its authors. Use permitted under Creative
Commons License Attribution 4.0 International (CC BY 4.0).
Another problem is low rating of completing courses. One of solutions is adaptive
learning. When learning is adaptive, every student has his own learning path
with his own preferences and features. Adaptive learning can make electronical
education more e ective [
Blockchain technology is a popular research topic. It is a distributed database which based on peer-to-peer network. Every transaction in the net must be conrmed with all the nodes. Every new con rmed transaction needs to be recorded to every node of the network. All the con rmed data is packed in special blocks; every block has a reference to previous block; so, there is a chain of blocks as all data storage.
Blockchain technology has some advantages of use in many areas like medicine,
realty and education [
Initially, blockchain was created for operation with cryptocurrency, but later it became independent platform for open and reliable data change in some peerto-peer network. Bitcoin blockchain is for cryptocurrency operations only, while Ethereum blockchain allows various transactions using smart contracts mechanism. A smart contract is a program, which describes some operations in the Ethereum blockchain, Operations become automated transactions between the nodes in the network. It allows to use blockchain in electronical education.
EduCTX project was developed from the Ethereum platform. EduCTX
provides internal tokens, which acts like virtual educational currency. Every student
has his own electronical ectx-wallet, which appends while learning process [
Nicosia University uses blockchain for certi cates' digital signature. Also, it
runs blockchain courses and accepts bitcoin as a payment currency [
Now there are some studies about certi cates' veri cation applying blockchain.
Con rmation records are automatically stored in the blockchain [
One of the most useful blockchain features is its ability to register and store all the events in the system. All these events are available for future processing.
The paper considers two cases of applying blockchain in the e-learning. 2 2.1
The rst case is generating reports. Today there are only determined options
for saving and reporting events in existing LMS [
So, grade report is a matrix: where every element is a dataset obtained from the function: (1) (2) (3) (4)
Function means program code, which are necessary to create for processing all the parameters.
When using blockchain, students form peer-to-peer network, so all their interactions are recorded into blockchain. Time parameters and users' identi ers are registered automatically by blockchain; therefore, grade report is matrix: where every element is a dataset obtained from the function:
U = fui;j g; ui;j = f (ai; ki; j; ti;j )
U 0 = fu0i;j g; u0i;j = f (ki;j )
Any report in every LMS can be considered similarly. Every report uses time parameters and users' identi ers. Blockchain makes reports more detailed and universal and doesn't require any data scheme. 2.2
Another case of using blockchain in education is student modelling in adaptive
learning. The student model is the central concept of adaptive learning [
If D is mentioned before input data for processing in the student modelling, then output dataset for the student model is a result of function: where A stands for an array of students' identi ers, like in the previous case; T stands for an array of all students' actions' timestamps, like in the previous case; f (D) = f (A; T; K); (5) K stands for all the other parameters, required for modelling including grades, actions, uploaded les, comments, viewed pages etc.
When using blockchain, every timestamp and user identi er are registered automatically, similarly to the previous case, so student model is a function: f 0(D) = f 0(K): (6)
So, blockchain using allows partially automate student modelling and makes input data more detailed. Also, no matter what type of adaptive model us used. Similar to previous case, no special data scheme is required. Blockchain is a universal tool for heterogeneous data registering and storing. Therefore, di erent models can be designed and implemented. 2.3
Both cases can be described in one model (see Fig. 1, Fig. 2).
It also can be presented mathematically. The rst case describes existing LMS:
where U means output dataset to form report or student's model, A means an array of users' identi ers, T means an array of all user actions' timestamps, main data K contains all the other user actions' parameters of learning process including grades, actions, uploaded les, comments, viewed pages etc.
Blockchain-based system simpli es the process:
Now, system has to register only K to generate U 0, because arrays A and T are registered by blockchain automatically.
It is important to say, that in both cases no special data scheme or structure is required. Any parameter can be added or selected for processing during system runtime.
The blockchain supports several properties that are considered in both cases: 1. Traceability. It can be easily tracked how an assessment was formed using a detailed report of all student's actions 2. Security. Data cannot be distorted. 3. Reliability. It can be found out that the assessment was obtained precisely by the methods provided for in the curriculum.
U 0 = f (K) (8)
Both models were implemented in a developer's version of the Ethereum blockchain using the smart contract mechanism. As it was said before, Ethereum is blockchain application platform, which allows to develop blockchain-based applications.
Input data was obtained from the educational center of Design in ITMO University. The system is based on the Moodle platform. At the moment, there are running more than 20 courses. Webmaster o ered three log les with results of studying \Design and layout using CorelDraw". Since the course started, there were about 270 students studied and their activities were recorded into three logs. The les contained following data: { First and last login of all users in the system, including date and time. { All students' grades. The students had to complete 11 home tasks and the nal test. Home tasks had to be loaded in the form of a le and were accompanied by a grade and teacher's comment. The grades were presented in a ve-point scale, as well as a percentage of the maximum possible value. The nal test consisted of 30 questions. Test grades were presented as a number of correctly answered questions, as well as a percentage of the maximum possible value. { All students' actions with date and time. The actions contained data on the following events: a student enrolling in a course, a student uploading a le, giving grades, test attempts, viewing a course section, as well as some system events like assigning a role to a user.
Every le had its own structure. So, a lot of heterogeneous data was available for processing with blockchain. There were more than 200 records with grades and about 40,000 actions.
This data was analyzed for types of events with their parameters and frequency. It is not necessary to model all the events from the input data. This paper needs only the most common types of students' actions: { task solving with grade and number of attempts; { page viewing; { le uploading with teacher's and student's comments.
Smart contract was created with Solidity language and Remix online complier. It contained event descriptions with a set of parameters for every event type.
There was a handler method for every event. Methods initiated events and recorded them into blockchain with event timestamp and author's address.
The contract was deployed and mined in the \dev" Ethereum network, available through \geth" console.
Event data was available through the \geth" console too (see Fig. 3, Fig. 4).
Dataset was exported into text le. It allows us to create any student's model and any educational report (see Fig. 5). User
Educational event
LMS
Smart-contract Calling smart-contract method
The output is a log exported into a .txt le. Here is a excerpt from the log with four entries of two types. The rst type is task assessment. It contains the "grade" parameter with a student grade as a percentage of the maximum value and the "task" parameter with an integer task number.
The second type is a student's action. It contains the "action" eld for the event's name.
Both types of records also have two parameters. "Time" is an integer that the blockchain registers. This is the time of adding the new block, which contains the transaction with this event. The "owner" parameter is the address of the node in the blockchain that triggered the event. This is a hexadecimal number. It serves as the user identi er in this dataset.
The di erence between input and output data can be presented in the table.
As it was shown, smart contract logged heterogeneous data of learning events and students' actions to the blockchain without using special data structures. The blockchain stored data in a uni ed registry and stores event timestamp and event author's address by itself.
As a result of the study we can say, that blockchain technology is a possibly good tool for electronic massive education. The next research task is to measure advantages of blockchain's implementation. The next practical task is to expand the system to the full version, including user interface, and also estimate the cost of using this model.