Kadri Umbleja

New concepts of automatic answer evaluation in competence based learning

This paper introduces current work with automatic evaluation that has been done to enable algorithms generating proper feedback according to the mistakes students have made. Learning environment using novel competence based approach has been used to implement those concepts and over the years data from different development steps has been collected that can be used to verify the benefits to students results of offered algorithm that mimics step-by-step student answering process.

Competence-based approach to learning

This paper describes competence based approach to complete learning process. The learning process is fully web-based and can be completed without attending campus. It consists of small exercises and lab experiments. Automatic evaluation by stimulating student answering process is explained. Labs are supported by HomeLabKit, small box that contains everything needed to perform lab tasks and can be lent from university. After two years of using competence-based learning in real learning, a lot of information has been collected that can now be analyzed.

Can K-12 Students Learn How to Program with just Two Hours?

This paper presents results of analyzing almost 1500 students’, from ages 6 to 18, feedback and test results to a two hour long programming workshop using LEGO Mindstorms kits. Students had very little or no previous programming experience. At the end of the workshop they were asked how the workshop affected their interest towards engineering in general, robotics and programming. They were also asked list of theoretical question about concepts and ideas learned during the workshop. Students’ answered were analyzed according to genre, age and school type. As a result, it can be concluded that two hours of intense programming workshop gives students similar basic understanding about programming concepts as other students with previous experience have. Primary school students were most excited about the workshop and stated highest positive change as a result of the workshop. On the other hand, they lacked ability to understand programming concepts. Therefore students between ages 11–12 and 16–18 benefit most from the workshop as they have high interest and they also are able to comprehend theory. Type of school student attends does not affect the results. Girls have lower previous interest towards three considered fields before the workshop and they have slightly lower results than boys.

Two-Dimensional Knowledge Model for Learning Control and Competence Mapping

The paper presents two-dimensional model for knowledge representation with volume as one variable and ability as another one. This makes possible describing current state of learner’s abilities and integration for higher level parameters e.g. grading related to course or other entities. Both values are related to atomized knowledge elements (competences) with volume interpreted as credit units and ability levels are formed during learning with application of forgetting. This model makes possible characterization (grading) of knowledge based on real abilities independently of predeclared courses and for ‘drop-outs’. So, on that bases one can obtain grade for some course if proper knowledge has been obtained in different courses and schools even when courses had not passed. Also this model helps to build connections between courses as using courses in the role of prerequisites becomes less usable. Not wasting knowledge obtained in MOOCs is another example with high drop-out levels where classical passed-failed model does not work.

Implementation of the robot arm in the interactive learning environment

While the robotics is becoming the more important branch of technology and industry, the teaching of industrial robotics has remained rather limited and expensive due to high cost of the industrial equipment and workrooms. Those drawbacks may be overcome by using the reduced size educational robot arm and the remote control of it via the interactive e-learning environment. The present paper describes the technical solutions that were designed and realized in order to give the possibility to students to learn the basics of industrial robot programming via the interactive environment.

Learning to Program with Lego Mindstorms – Difference Between K-12 Students and Adults

In this paper, a workshop that was originally designed for K-12 students as a first introduction into programming by using Lego Mindstorms is conducted on adults coming from different companies. The aim of the study is to compare attitudes and results of adult learners with K-12 students and to see how their interests and understanding of programming concepts differ from the original audience. It is found that the exact same workshop is suitable for older audience and, as expected, the adults are showing slightly better learning outcomes. Using Lego as a tool for teaching is not a problem for adults.

Predicting Students’ Behavior During an E-Learning Course Using Data Mining

This paper introduces a process of building a prediction model for student’s final grade and time of finishing, based on students’ previous behavior. Prediction model was developed using data mining with regression analysis, principle component analysis and hierarchical clustering of symbolic histogram valued data. 35 different features of students’ activates was considered but only the 9 most important, so called principle components, were used in the model. Then, using histogram valued data - a type of symbolic data that allows learning processes to be described in a more natural form, and a hierarchical clustering, previous students’ behaviors were grouped. For an accurate prediction, a closest cluster to student’s current progress was found. To verify the model’s correctness, predictions were tested on a largest course in e-learning system in 2015 fall semester. The model was found to work sufficiently.

Similarity and Dissimilarity Measures for Mixed Feature-type Symbolic Data

This paper presents preliminary results for the similarity and dissimilaritymeasures based on the Cartesian System Model (CSM) that is a mathematical modelto manipulate mixed feature-type symbolic data. We define the notion of conceptsize for the description of each object in the feature space. By extending the notionto the concept sizes of the Cartesian join and the Cartesian meet of the descriptionsfor objects, we can obtain various similarity and dissimilarity measures. We presentespecially the asymmetric similarity measure, and the symmetric similarity anddissimilarity measures useful for pattern recognition problems.

Competence based interactive learning with HomeLabKits: Experience and work in progress

This paper describes the implementation and the application experience of learning with the Home Laboratory Kits. HomeLabKit is a small suitcase containing everything needed to perform laboratory exercises in the courses of the electrical engineering. The HomeLabKit has been tested in the real learning process for several years.

The method of quantum optical communication based on entangled photon pairs

During the last two decades have been successfully developed the quantum cryptography techniques that rely either on manipulation of the polarized single photons or the entangled photon pairs. At that has been found confirmation the fact the quantum correlations between the entangled photons are nearly or totally instantaneous but those correlations cannot be used for the direct superluminal information transfer as the state of the photon in the pair is unknown before the first measurement. Technology of entangled photons has made possible the modified double slit experiments where the appearance of the interference picture in detecting of the “signal” photons may be controlled by the manipulation of the spatially separated “idle” photons. In the present paper we discuss the possible superfast communication channel between the entanglement-linked distant parties using the remote manipulation of the interference picture formed statistically by the number of signal photons.