Search

The blended learning (BL) services presented for the academic accreditation process of universities and other educational institutions is a policy that has been used rapidly in the educational area. This process represents a real challenge and a need for institutions to function by the challenges in the ways of development for the attainment of faculty and student satisfaction. The need for learning management systems (LMS) becomes a real advance to control the management of technological development. The purpose of this study is to overcome these challenges and speed up the whole e-learning process in a different manner. The study revealed the existence of an important gap between the important features of blended learning applications and the control management of LMS. The finding shows a systematic analysis of existing technologies and their effects on accreditation checkpoints approval for the underlying conception of the assessment processes involve perspectives that relate to what is now considered valuable or desirable concerning the institution of higher education.

Principal Component Analysis (PCA) is a commonly used unsupervised exploratory analysis technique. It is also frequently used in dimensionality reduction. This preliminary paper investigates the feasibility of three variants of PCA, i.e. independent PCA (iPCA), sparse PCA (sPCA), and sparse independent PCA (siPCA) on forensic classification of paper based on their IR spectral data. After that, Linear Discriminant Analysis (LDA) models were built using the Principal Components (PCs) produced by the PCA and the three aforementioned variants. The performances of all these four LDA models, i.e. PCA-DA, iPCA-DA, sPCA-DA and siPCA-DA, were evaluated via leave-one-out cross-validation on the data set. The results obtained show that iPCA-DA and siPCA-DA are the most effective models with 100.0% classification accuracy. Then, the effectiveness of siPCA and iPCA models was evaluated based on posterior probability used for predictions of class membership that were derived from leave-one-out cross-validation. As a conclusion, siPCA is identified as the best classification model.

The concept of this paper focuses on the area of testing university students while using tools of artificial intelligence. The sample set of students were students of the third year of bachelor studies of Applied Informatics and the test was focused on testing English: grammar, vocabulary, IT English. The introductory part describes current possible ways of testing students, mainly in paper- or computer-based tests. The paper-based method is the traditional one while the computer-based one can take advantage of modern tools of artificial intelligence, such as expert systems and neural networks. The core of this paper lies in the comparison of the results that are acquired from paper-based and computer-based testing which uses artificial intelligence tools. The final part envisages the most optimal ways of testing university students.

Historians and librarians are interested in watermarks and mould surface patterns in historic papers, because they represent the “fingerprints” of antique papers. However, these features are usually covered or hidden by printing, writing or other media. Different techniques have been developed to extract the watermarks in the paper while avoiding interference from media on the paper. Beta radiography provides good results, but this method cannot be widely used because of radiation safety regulations and the long exposure times required due to weak isotope sources employed. In this work, two promising methods are compared which can be used to extract digital high-resolution images for paper watermarks and these are electron radiography and low energy X-ray radiography. For electron radiography a “sandwich” of a lead sheet, the paper object, and a film in a dark cassette, is formed and it is exposed at higher X-ray potentials (> 300 kV). The photoelectrons escaping from the lead sheet penetrate the paper and expose the film. After development, the film captures the watermark and mould surface pattern Images for the paper being investigated. These images are then digitized using an X-ray film digitizer. The film employed could potentially be replaced by a special type of imaging plate with a very thin protection layer to directly generate digital Images using computed radiography (CR). For the second method, a low energy X-ray source is used with the specimen paper placed on a digital detector array (DDA). This method directly generates a low energy digital radiography (DR) image. Both methods provide high quality images without interference from the printing media, and provide the potential to generate a “fingerprint” database for historical papers. There were nevertheless found to be differences in the images obtained using the two methods. The second method, using a low energy X-ray source, has the potential to be integrated in a portable device with a small footprint incorporating user safety requirements. Differences obtained using the two methods are shown and discussed.

EMG is a medical technique that measures muscle response or electrical activity in response to a nerve’s stimulation of the muscle [10]. In this paper we have classified multiple hand gestures using small number of electrodes. This report aims to review current use of EMG based control system and explain the existing challenges. The stages involved includes data acquisition, feature extraction, classification and controller. Some of the application includes Arm rehabilitation, game controller, wheel chair controllers, etc.

Several approaches have been proposed to help researchers in retrieving relevant and useful information from the cyber-ocean of information. However, the approaches assumed the academic content to be easily accessible, which is not always the case considering the copyright restrictions. Different from the existing works, we proposed a collaborative approach that leverages public contextual metadata to personalize scholarly recommendations. Our proposed approach has the ability to recommend research papers to the individual researchers regardless of the researcher expertise and research field. As demonstrated using a freely accessible dataset, our proposed approach have shown significant improvements over another baseline method.

An algorithm for constructing super-large semantic networks has been developed in current work. Algorithm was tested using the “Cosmos” category of the Internet encyclopedia “Wikipedia” as an example. During the implementation, a parser for the syntax analysis of Wikipedia pages was developed. A graph based on list of articles and categories was formed. On the basis of the obtained graph analysis, algorithms for finding domains of high connectivity in a graph were proposed and tested. Algorithms for constructing a domain based on the number of links and the number of articles in the current subject area is considered. The shortcomings of these algorithms are shown and explained, an algorithm is developed on their joint use. The possibility of applying a combined algorithm for obtaining the final domain is shown. The problem of instability of the received domain was discovered when starting an algorithm from two neighboring vertices related to the domain.