Sustav preporučivanja aktivnosti za računalom podržano suradničko učenje

doctoral thesis

• Title (english): Recommender system for activities in computer-supported collaborative learning
• Author: Martina Holenko Dlab; MBZ: 298604
• Mentor: Vedran Mornar (mentor)
• Mentor: Nataša Hoić-Božić (komentor); MBZ: 199562
• Committee member: Vedran Mornar (predsjednik povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; (Department of Applied Computing); Zagreb
• Defense date and country: 2014, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing; Program Engineering
• Universal decimal classification (UDC): 621.3 - Electrical engineering

Abstract

Sustavi preporučivanja pružaju podršku korisnicima prilikom pristupa korisnim ili zanimljivim sadržajima. U procesu preporučivanja uvažavaju se individualne karakteristike korisnika, čime se osigurava personalizacija. Tehnike preporučivanja se koriste i u domeni e-učenja, no kod većine postojećih sustava skup sadržaja koji se preporučuju je ograničen na nastavne materijale i kolegije u cjelini. Kod suvremenog e-učenja koje pretpostavlja izvođenje suradničkih aktivnosti (e-aktivnosti) različitim... More alatima postoje i drugi aspekti koje je moguće prilagoditi karakteristikama studenata. Preporuke se tako mogu odnositi na odgovarajuće suradnike ili alate za e-aktivnost, savjete kojima će se podržati studenta u njenoj realizaciji, ali i na e-aktivnost u cjelini. U doktorskom radu je opisan vlastiti model obrazovnog sustava preporučivanja za personalizaciju izvođenja e-aktivnosti koje se realiziraju uz pomoć alata Weba 2.0. Preporuke su namijenjene studentima i grupama te uključuju četiri vrste sadržaja: izborne e-aktivnosti, suradnike, alate Weba 2.0 i savjete za povećanje razine aktivnosti. Glavne komponente opisanog sustava su model aktivnosti, podsustav za modeliranje korisnika i podsustav za određivanje preporuka. Modelom aktivnosti se predstavlja planirani tijek aktivnosti e-učenja i sadržaji za preporučivanje. Podsustav za modeliranje sadrži komponentu za prikupljanje podataka o akcijama koje studenti izvršavaju pomoću alata Weba 2.0 te modele studenta i grupe kojima se predstavlja sljedeće karakteristike: preferencije stilova učenja, preferencije alata Weba 2.0, razine znanja i razine aktivnosti. Preporuke se određuju u podsustavu za određivanje preporuka primjenom algoritama koji uključuju pedagoška pravila. Nastavnik modificira navedena pravila u skladu sa željenim pedagoškim pristupom te time upravlja procesom preporučivanja. Opisani model sustava implementiran je kao obrazovni sustav preporučivanja nazvan ELARS (E-Learning Activities Recommender System). Prilikom vrednovanja sustava naglasak je bio na vrednovanju s pedagoškog aspekta. Vrednovanje je izvršeno u nastavnom radu sa studentima pri čemu se utvrdila djelotvornost korištenja sustava u e-kolegijima. Less

Abstract (english)

Recommender systems provide support in accessing useful or interesting content. In the recommendation process individual characteristics of the user are taken into account in order to provide personalization. Recommendation techniques are also used in the domain of e-learning, but in the most of existing systems set of items that is recommended is limited to set content to learning materials and courses in general. In the e-learning which includes collaborative learning activities (e-tivities)... More different aspects can be adapted to students' characteristics. Thus, recommendations can also refer to the appropriate collaborator or tool for e-tivity, advice that will support students in its implementation, or e-tivity as a whole. The dissertation presents own model of the recommender system that enables personalization within a structured sequence of collaborative e-tivities realized using Web 2.0 tools. The system contains the following components: the activity model, the subsystem for user modeling and the subsystem for determining recommendations. The activity model represents course activities and items that can be recommended. Those are: optional e-tivities, collaborators, Web 2.0 tools and advice. The subsystem for user modeling represent student's and group's characteristics that are used for personalization, and includes the original procedure for activity level estimation for students and groups. The activity level represents quantity and continuity of student's (group's) contributions in e-tivity relatively to other's contributions. The data is calculated based on certainty factors theory and, together with other user characteristics, used in the recommendation process. Recommendation process is carried out within the subsystem for determining the recommendations which contains pedagogical rules. Teachers modify existing rules in order to define the recommendation criteria. The prototype of the system was built and called ELARS – E-learning activities recommend system. The proposed methods, algorithms and pedagogical rules were implemented within ELARS system and verified in the actual environment. The results of the evaluation indicated that the system is effective in the sense that the students who use it achieve better results on e-courses. Additionally, the results of a questionnaire showed that the students find the system useful for e-tivities, that they are satisfied with its interface and received recommendations. Less

• Keywords: sustav preporučivanja; e-učenje; računalom podržano suradničko učenje; personalizacija; dizajn učenja; izborne e-aktivnosti; alati Weba 2.0; razina aktivnosti; savjeti
• Keywords (english): recommender system; e-learning; computer-supported collaborative learning; personalization; learning design; optional e-tivities; Web 2.0 tools; activity level; advice
• Language: croatian
• URN:NBN: urn:nbn:hr:168:848559
• Study programme: Title: Electrical Engineering and Computing; Study programme type: university; Study level: postgraduate; Academic / professional title: Doktor znanosti elektrotehnike i računarstva (Doktor znanosti elektrotehnike i računarstva)
• Type of resource: Text
• Extent: 181 str. ; 30 cm.
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-04-18 14:47:39