|
SITAC – Innovative Computerized Adaptive Testing System Servet Giafer, Cristina Giafer |
1-20 |
|
Usability Evaluation of Gesture-Based Interactions using a Smart Watch Ioana-Crinela Potinteu, Teodor Ștefănuț |
21-38 |
|
Computer-based Story Generation. An Analysis from a Phenomenological Standpoint Ştefan Trăuşan-Matu |
39-53 |
|
Extracting human features to enhance the user experience on a training station for manual operations Alin-Marius Cruceat, Alexandru Matei, Bogdan-Constantin Pîrvu, Alexandru Butean |
54-66 |
S.C. Soft Business Union S.R.L.
6 Blv Regina Maria, sector 4, 040125, Bucharest, Romania
Abstract. The computerized adaptive testing is an approach of the differential assessment which adapts the questions that are asked to the candidate’s ability level. Thus, the computer selects and displays the questions, then records and processes the candidate’s answers. The items’ selection is adaptive and it depends, on one hand, on the candidate’s answers to the previous questions and on the statistical qualities of the items, on the other hand. As compared to the traditional testing methods, where all candidates receive the same items, the computerized adaptive testing is managing a higher percentage of items, with appropriate difficulty levels. The items’ adaptive selection process results in higher levels of the scores’ precision and shorter tests. SITAC - the Innovative Computerized Adaptive Testing System represents a complex and innovative software product based on a computerized adaptive testing platform through which the testing of the persons will be performed, for the purpose of estimating their abilities, based on the answer previously delivered.
Keywords: Computerized Adaptive Testing, Item Response Theory, Item Characteristic Curve, Logistic Model.
Cite this paper as:
Giafer, S., Giafer, C. SITAC – Innovative Computerized Adaptive Testing System.
International Journal of User-System Interaction 12(1),
1-20, 2019.
Technical University of Cluj-Napoca, Computer Science Department
26-28 G. Baritiu, 400027, Cluj-Napoca, Romania
Abstract. This paper describes the testing methodology applied and the results obtained in evaluating the usability of a system for gesture recognition based on a smartwatch and a mobile phone. The data acquisition is performed using a single accelerometer with 3 axes, using an approach based on button segmentation. For our testing purposes, the system has been set up to recognize 4 arm gestures, that represent specific commands for a music player that is running on the mobile phone. This setup has also offered us the means to provide users with easy to recognize feedback on how their actions have been interpreted by the system. Inspired from real-life scenarios, our testing methodology has considered three major contexts in which the users should be able to easily interact with the system: while sitting, when walking and while running.
Keywords: human-computer interaction, accelerometer, gesture recognition, smartwatch, classification, dynamic time warping.
Cite this paper as:
Potinteu, I.-C., Ștefănuț, T. Usability Evaluation of Gesture-Based Interactions using a Smart Watch.
International Journal of User-System Interaction 12(1),
21-38, 2019.
1 University Politehnica of Bucharest
313 Splaiul Independentei, Bucharest, Romania
2 Institutul de Cercetări în Inteligenţa Artificială
Calea 13 Septembrie nr. 13, Bucureşti
3 Academy of Romanian Scientists
Splaiul Indpendentei 54, Bucharest, Romania
Abstract. This paper has a starting point on the idea that we need a phenomenological, post- structuralist perspective for generating realistic narratives with Natural Language Processing techniques. Several classifications of existing Natural Language Generation systems have been proposed. A critical analysis is done of the existing theories and approaches for generating narratives. Requirements of realistic narratives are discussed and ideas of using chronotopes and polyphony for these aims are introduced.
Keywords: story generation, natural language processing, natural language generation, polyphonic model, chronotopes, artificial intelligence, creative writing, phenomenology.
Cite this paper as:
Trăuşan-Matu, S. Computer-based Story Generation. An Analysis from a Phenomenological Standpoint.
International Journal of User-System Interaction 12(1),
39-53, 2019.
1 Lucian Blaga University of Sibiu, Connected Intelligence Research Center
10 Victoriei Boulevard, Sibiu, Romania
2 Lucian Blaga University of Sibiu, Department of Computer Science
10 Victoriei Boulevard, Sibiu, Romania
Abstract. This paper presents an image analysis approach for extracting human features. These features are used in a physical training station for manual operations to support training environment adaptation based on predefined user profiles. A Kinect for Microsoft Xbox provides the RGBD data flow that is used to process the user features utilized furthermore for adjusting a training table and thus supporting the user in executing his training routine in an ergonomic and efficient manner. The development process and the integration of the training station with other smart devices will rely on these human features and specific user profiles to generate adapted training instructions.
Keywords: Training systems, motion training systems, image processing, feature extraction, visual feedback, Kinect for Microsoft Xbox 360.
Cite this paper as:
Cruceat, A.-M., Matei, A., Pîrvu, B.-C., Butean, A. Extracting human features to enhance the user experience on a training station for manual operations.
International Journal of User-System Interaction 12(1),
54-66, 2019.