Romanian Journal of
Human - Computer InteractionVol.8, No.3, 2015
ISSN 1843-4460
Contents
Ranking of comments published on YouTube using natural language processing techniques
Iulian Radu, Traian Rebedea
173 - 194 The educational potential of Facebook use by students in two Romanian universities
Valentina Iuliana Manea, Gabriel Gorghiu, Dragoş Daniel Iordache
195 - 208 Accessibility issues in Romanian municipal websites
Maria-Gheorghe Moisii
209 - 222 Evaluation of collaborative learning in chats, based on the analysis of repetitions and altruism
Ştefan Trăuşan-Matu, Roxana-Ioana Murăruş
223 - 236 Visual analytics on multidimensional massive data
Dorian Gorgan, Mircea Cǎtǎlin Cǎtanǎ, Teodor Ştefănuţ
237 - 255
Abstracts
Ranking of comments published on YouTube using natural language processing techniques
Iulian Radu, Traian Rebedea
University Politehnica of Bucharest, Faculty of Automatic Control and Computers,
313 Splaiul Independenţei , 060042 Bucharest, Romania
E-mail: iuliradu@gmail.com, traian.rebedea@cs.pub.ro
Abstract. Publishing of comments is a form of interaction that has become an important part of the Social Web phenomenon. Comments are enhancing online interaction, however their visualization and ranking becomes extremely difficult or even impossible for videos or other types of online articles that have millions of comments linked to them. This paper presents a method to identify the most relevant comments published for a video uploaded on the YouTube video-sharing platform, offering at the same time a detailed analysis of the features that are relevant for the most important comments linked to an article. The analysis we have performed shows that videos from different categories have different relevant attributes and there are important differences between these attributes for specific pairs of categories (e.g. Music vs. Education).
Keywords: information retrieval, natural language processing, social web, comments analysis, relevance computation, ranking.
Cite this paper as:
Radu, I., Rebedea, T. Ranking of comments published on YouTube using natural language processing techniques. Revista Romana de Interactiune Om-Calculator 8(3), 173-194, 2015.
The educational potential of Facebook use by students in two Romanian universities
Valentina Iuliana Manea1, Gabriel Gorghiu2, Dragos Daniel Iordache3
1Universitatea Tehnică de Construcţii Bucureşti
Bd. Lacul Tei, Nr. 124, Bucureşti
2Universitatea Valahia din Târgovişte
Bd. Carol I, Nr. 2, 130024, Târgovişte
3Institutul Naţional de Cercetare-Dezvoltare în Informatică – ICI Bucureşti
Bd. Mareşal Averescu, Nr. 8-10, 011455, Bucureşti
E-mail: valentinamanea@yahoo.com, ggorghiu@yahoo.com, iordache@ici.ro
Abstract. Facebook is the most popular social networking website. This popularity has led Facebook to be an important subject of research, especially regarding the impact of using among young people. This article presents an analysis of the characteristics of Facebook use by students at two Romanian universities: Technical University of Civil Engineering and University Valahia of Targoviste. The results show marginally significant differences as regarding the network size and intensity of use between men and women. Female students have a greater number of friends and spend more time on Facebook than male students. Regarding the motives to use, students are primarily interested to use Facebook to communicate with friends, secondly for keeping in touch with former colleagues in high school and a thirdly, to find out what is new. Regarding the educational usefulness, students consider that Facebook is firstly useful for collaboration, secondly for information and thirdly for active participation.
Keywords: Facebook, social networking websites, educational use of Facebook, gender differences, motives of using Facebook, Facebook usefulness.
Cite this paper as:
Manea, V.I., Gorghiu, G., Iordache, D.D. The educational potential of Facebook use by students in two Romanian universities. Revista Romana de Interactiune Om-Calculator 8(3), 195-208, 2015.
Accessibility issues in Romanian municipal websites
Maria Gheorghe-Moisii
National Institute for Research and Development in Informatics - ICI Bucharest
Bd. Averescu, Nr. 8-10, 011455, Bucharest
E-mail: moise@ici.ro
Abstract. The accessibility of public websites is essential for the successful implementation of the information society. Given that the society includes people with various forms of disability (visual impairments, cognitive, hearing, etc.) they should also have equal opportunities in accessing and using the public information. This paper presents the results of a study carried on in 2015 that assess the accessibility of 150 municipal web sites. The evaluation was done with Total Validator. The accessibility issues have been compared with the results of previous studies. The results show a low level of web accessibility and instability of results in time.
Keywords: web accessibility, online services, WCAG 2.0, website evaluation.
Cite this paper as:
Gheorghe-Moisii, M. Accessibility issues in Romanian municipal websites. Revista Romana de Interactiune Om-Calculator 8(3), 209-222, 2015.
Evaluation of Collaborative Learning in Chats, Based on the Analysis of Repetitions and Altruism
Ştefan Trăuşan-Matu1,2,3, Roxana-Ioana Murăruş1
1Universitatea Politehnica din Bucureşti
Splaiul Independenţei nr. 313, Bucureşti,
E-mail: stefan.trausan@cs.pub.ro, murarus.roxana@gmail.com
2Institutul de Cercetări în Inteligenţa Artificială
Calea 13 Septembrie nr. 13, Bucureşti
3Academia Oamenilor de Ştiinţă din România
Splaiul Independenţei nr. 54, Bucureşti
Abstract. The paper presents a system for the evaluation of the students participating to a computer supported collaborative learning session using chat in English language. To each participant was assigned a different concept to sustain and they have to debate on them. The analysis is based on the usage of repetitions, focusing on the functions they may have, from socialization to emphasizing an idea. Natural Language Processing techniques are used. The number and type of words repeated by each participant is a measure of the implication of that student in each discussion thread. The evaluation of the students is done by providing one of the variants: Very good, Good, Satisfying, and Unsatisfying, starting from the number of words specific to debates, of those of the domain of discussion, on the length of the words, and from a measure of the implication in the discussion, considering the personal and the others’ topics. The percentage in which each of these factors is considered may be changed by the user through the interface.
Keywords: polyphony, dialogism, repetitions, CSCL, suffix tree, altruism.
Cite this paper as:
Trausan-Matu, S., Murarus, R.I. Evaluation of Collaborative Learning in Chats, Based on the Analysis of Repetitions and Altruism. Revista Romana de Interactiune Om-Calculator 8(3), 223-236, 2015.
Visual Analytics on Multidimensional Massive Data
Dorian Gorgan, Mircea Cătălin Cătană, Teodor Ştefănuţ
Computer Science Department, Technical University of Cluj-Napoca
Str. G. Bariţiu, Nr. 28, 400027, Cluj-Napoca
E-mail: {dorian.gorgan, teodor.stefanut}@cs.utcluj.ro, mircea.catalin.catana@gmail.com
Abstract. The paper proposes visual analysis techniques as a solution to the scientific community concerned with developing performant algorithms for exploring and analyzing massive and multidimensional data. The actually high performance computation infrastructures such as cloud, grids, multicore, and clusters, are able more or less to cover the computation requirements for transformation, classification, and highlighting of significant data. Even so, the analytical capacity of these systems is still limited. Nevertheless the human brain has a much greater analytical and synthetic capacity. This paper presents the development and the experiments on the visual analyzing techniques in the domain of strategic games. The visual techniques of developing strategies for the tower defense games are exemplified and evaluated. The metrics for strategy evaluation and the experiments on interactive visual navigation techniques throughout the value space in order to support the convergence toward the optimum solutions are detailed. The visual analyzing techniques and the exemplified solutions could be of interest for players who wish to improve their performance and for strategic game developers as well.
Keywords: visual analysis, interactive navigation, value space, game strategy.
Cite this paper as:
Gorgan, D., Catana, M.C., Stefanut, T. Visual Analytics on Multidimensional Massive Data. Revista Romana de Interactiune Om-Calculator 8(3), 237-255, 2015.