Journal of Software Engineering & Intelligent Systems

ISSN: 2518-8739 (Online)
APRIL 2021 | VOLUME. 6  ISSUE. 1
Title:

EVALUATION MODEL FOR THE USABILITY OF WEB-BASED LEARNING MANAGEMENT SYSTEMS WITH USER PROFILE

Authors:

Ghanim Hussein Ali Ahmed, Laszlo Kovacs, Gfary Hassan Hajhamed

Abstract:

The main purpose of Web-Based Educational Resources Management Systems (WERMSs) is to deliver knowledge, share information to help learners in their learning activities with an effective and efficient way by involving advanced electronic technologies. However, the Usability of these systems that is the degree of these systems to enable their users to use them effectively, efficiently and with satisfaction in a specified context of use, is one the challenges faced in the design of these systems. One of the main purposes of Human Computer Interaction (HCI) is using usability concept to create WERMSs that provide a relationship between user information and the WERMSs-content. Consequently, the next step is to evaluate the WERMSs will be based on usability metrics. This paper proposes a model for usability of WERMSs. The model introduced effectiveness, efficiency satisfaction, learnability, interactivity, consistency, motivation, learner's control, and user profile (age, gender, and level of experience) as the attributes that determine the usability of such systems. The model tested and verified using questionnaires and experiments. The results showed up that effectiveness, efficiency, satisfaction, learnability, motivation, interactivity, consistency, learner's control, and user profile (age, gender, and level of experience) affects the usability of WERMSs.

Keywords: web-based applications; e-learning management systems; human computer interaction; usability evaluation; e-learning;
Available: EVALUATION MODEL FOR THE USABILITY OF WEB-BASED LEARNING MANAGEMENT SYSTEMS WITH USER PROFILE by Ghanim Hussein Ali Ahmed, Laszlo Kovacs, Gfary Hassan Hajhamed is available under cc by-nc 4.0 License on 30th April 2021. Permissions beyond the scope of this license may be available at JSEIS.
Download PDF

Title:

A SPATIAL AND FREQUENCY BASED METHOD FOR MICRO FACIAL EXPRESSIONS RECOGNITION USING COLOR AND DEPTH IMAGES

Authors:

Seyed Muhammad Hossein Mousavi

Abstract:

Human face states the inner emotions, thoughts and physical disorders. These emotions are expressed on the face via facial muscles. The estimated time through which a facial expression occurs on the face is between 0.5 to 4 seconds, and a micro expression between 0.1 to 0.5 seconds. Obviously, for the purpose of recording micro expressions, obtaining videos frames between 30 up to 200 frames per second is essential. This research uses Kinect V.2 sensor to get the color and depth data in 30 fps. Depth image stores useful 2.5-Dimentional information from skin wrinkles which is the main key to recognize even slightest micro facial expressions. Experiment starts with splitting color and depth images into facial parts, and after applying preprocessing techniques, features extraction out of both type of data in spatial and frequency domain takes place. Some of the features which are used in this study are Histogram of Oriented Gradient (HOG), Gabor Filter, Speeded Up Robust Features (SURF), Local Phase Quantization (LPQ), Local Binary Pattern (LBP). Non dominated Sorting Genetic Algorithm II (NSGA-II) feature selection algorithm applies on extracted features to have faster learning process and finally selected features are sent to neuro-fuzzy and neural network classifiers. Proposed method is evaluated with the benchmark databases such as, Eurecom Kinect Face DB, VAP RGBD-T Face, JAFFE, Face Grabber DB, FEEDB, and CASME. Moreover, the proposed method is compared with other similar methods and Convolutional Neural Network (CNN) method on mentioned databases. The results are really satisfactory, and it indicates classification accuracy improvement of proposed method versus other methods.

Keywords: micro facial expressions recognition; Kinect sensor; depth data; spatial and frequency domain; evolutionary feature selection; neuro-fuzzy classifier;
Available: A SPATIAL AND FREQUENCY BASED METHOD FOR MICRO FACIAL EXPRESSIONS RECOGNITION USING COLOR AND DEPTH IMAGES by Seyed Muhammad Hossein Mousavi is available under cc by-nc 4.0 License on 30th April 2021. Permissions beyond the scope of this license may be available at JSEIS.
Download PDF

Title:

EXTENDED DYNAMIC SOFTWARE PRODUCT LINES ARCHITECTURES FOR CONTEXT INTEGRATION AND MANAGEMENT

Authors:

Amougou Ngoumou, Marcel Fouda Ndjodo

Abstract:

Nowadays, many embedded system families and application domains such as ecosystems, service-based applications, and self-adaptive systems in pervasive systems and cloud computing require runtime capabilities for flexible adaptation, reconfiguration, and post-deployment activities. However, we still have semi Dynamic Software Product Lines (DSPLs) architectures that need improvement for providing mechanisms for runtime adaptation and behaviour of products. There is an advancement toward designing more dynamic software architectures and building more adaptable software able to handle autonomous decision-making, according to varying conditions in the context. Recent development in DSPLs attempt to address the challenges of the dynamic conditions of such systems but the state of these solution architectures is still modest. In order to provide a more comprehensive architecture there is a need to take into account the context of the DSPLs models, their solution architectures and to cope with uncertainty at runtime. In this research work, we provide a formal representation of the context in the solution architecture of DSPLs models and the management of their behaviour according to that context.

Keywords: dynamic software product lines; self-adaptive software; runtime adaptation; solution architectures; context modelling; variability management;
Available: EXTENDED DYNAMIC SOFTWARE PRODUCT LINES ARCHITECTURES FOR CONTEXT INTEGRATION AND MANAGEMENT by Amougou Ngoumou, Marcel Fouda Ndjodo is available under cc by-nc 4.0 License on 30th April 2021. Permissions beyond the scope of this license may be available at JSEIS.
Download PDF

Title:

AN EXTENSION TO WOLF SHEEP PREDATION (DOCKED HYBRID) AGENT-BASED MODEL IN NETLOGO

Authors:

Muhammad Husnain, Numan Shafi

Abstract:

In old Wolf Sheep Predator and Prey model a relationship among two different types of models is shown in a relatively natural ecosystem using NetLogo. One of these is a simple mobile agent based model for simulation while other is an aggregate model. Evaluation of both the models can be done in one by one manner or by using parallel comparison button to see behavior of both the models for the same inputs. Agent based model also captures very basics of the reality to make simulation. In this model, a random procedure is used to simulate wandering of predator (Wolves) and Prey (Sheep) in specified area. A fixed gain in energy is awarded to the Wolf on each interaction with Prey while each step on which this type of interaction is not taking place, causes a constant decrease in energy. Thus, for wolves to survive it is necessary to eat sheep to regain their energies otherwise they will start dying due to lack of energy. This construct has an assumption that reproduce rate or birth rates of wolves and sheep have a constant probability and is proportional to the population of both species. Aggregate model is a construct that is modeled using system dynamics and only consists of standard LotkaVolterra equations of predator and prey. But our new extension to this old model introduced a third agent property Food (Grass) as a competitor to make it more realistic. Now the interaction of these agents has introduced more complexity in the model and also altered the equilibrium parameters. When sheep encounters a cell with grass it eats it only if his energy is not reached to its maximum limit. Each grassy cell adds in energy of the sheep. Model initially have the assumption that probability of reproduction time of grass at each cell is constant. After, we also have introduced a new docked model for comparison between newly created Wolf-Sheep agent based model and chained aggregated model. The comparison shown that for all equilibrium states actual Wolves in agent based model are 3 times lesser than in system dynamics aggregated model.

Keywords: predator-prey interactions; aggregated predator-prey model; agent based predator-prey model; NetLogo; docked hybrid simulation comparisons; system dynamics;
Available: AN EXTENSION TO WOLF SHEEP PREDATION (DOCKED HYBRID) AGENT-BASED MODEL IN NETLOGO by Muhammad Husnain, Numan Shafi is available under cc by-nc 4.0 License on 30th April 2021. Permissions beyond the scope of this license may be available at JSEIS.
Download PDF

Title:

IMPROVING THE KNN ALGORITHM BY USING WEIGHTED EUCLIDEAN DISTANCE

Authors:

SeyedSoroosh Azizi

Abstract:

The k-nearest neighbors (KNN) algorithm is one of the most popular machine learning algorithms and it performs very well on many machine learning problems. KNN is a nonparametric method used for both classification and regression problems. In this paper, we introduced a modified KNN method that outperforms KNN in both classification and regression settings. In the modified KNN method we use weighted Euclidean distance rather than simple Euclidean distance to find k-nearest neighbors of each observation. Both methods are tested on three datasets, one simulated and two real datasets, and results indicate that the modified KNN outperforms the KNN method in all cases.

Keywords: machine learning; KNN method; regression; classification; Euclidean distance; k-nearest neighbors;
Available: IMPROVING THE KNN ALGORITHM BY USING WEIGHTED EUCLIDEAN DISTANCE by SeyedSoroosh Azizi is available under cc by-nc 4.0 License on 30th April 2021. Permissions beyond the scope of this license may be available at JSEIS.
Download PDF

Archive

Volume 6
Issue 2   New
Issue 1
Volume 5
Issue 3
Issue 2
Issue 1
Volume 4
Issue 3
Issue 2
Issue 1
Volume 3
Issue 3
Issue 2
Issue 1
Volume 2
Issue 3
Issue 2
Issue 1
Volume 1
Issue 2
Issue 1
Journal PicsCaomei Publishers © 2016-2021
f t