A Model for Personalized Learning Through IDTV - CiteSeerX

A Model for Personalized Learning Through IDTV Marta Rey-López, Ana Fernández-Vilas, and Rebeca P. Díaz-Redondo Department of Telematic Engineering, University of Vigo, 36310, Spain {mrey, avilas, rebeca}@det.uvigo.es

Abstract. Interactive Digital TV (IDTV) opens new learning possibilities where new forms of education are needed. In this paper we explain a new conception of t-learning experiences where TV programs and learning contents are combined. In order for its creation to be possible we will use Adaptive Hypermedia techniques and Semantic Reasoning to design an Intelligent Tutoring System (ITS) whose tasks consist in selecting, combining and personalizing the contents to construct these learning experiences.

1 Introduction The arrival of IDTV makes the access to distance education easier, since about 98% of European homes have at least one television set, whereas the penetration of Internetenabled computers is lower than 60% [1]. Apart from wide-world usage, TV is considered by the viewer trustworthy in reference to broadcast content and easy to operate. These conditions are an ideal starting point for TV-based interactive learning, referred to as t-learning. In fact, education has always been present on TV, embedded in documentaries or programs for children—e.g. Sesame Street. To designate this form of entertainment designed to be educational, in 1973, Robert Heyman coined the term edutainment. Today, some TV channels have developed t-learning contents in this direction. In the UK, we can find some examples of games and interactive stories for children, as well as documentaries with additional contents, e.g. Walking with beasts produced by the BBC [1]. In Portugal, TV Cabo has developed several edutainment applications, some of them adapted from existing web sites, like Ciberdúvidas —resolving doubts regarding the Portuguese language [2]. Apart from introducing education into TV programs, transferring traditional structured courses to IDTV is also possible, using this medium solely as a means of transmission for education: transmitting on TV the image of the teacher to the students and vice versa [3] or broadcasting on TV typical e-learning courses, based on text and images [4]. One step further, the scenario for t-learning developed by our research group [5] has improved on these, since learning resources are designed especially for TV and so are based on audio and video content. However, the approaches mentioned above isolate learning elements from TV programs. The approach we propose looks for a new conception of learning through TV, 

Partly supported by the R+D project TSI 2004-03677 (Spanish Ministry of Education and Science) and by the EUREKA ITEA Project PASSEPARTOUT.

V. Wade, H. Ashman, and B. Smyth (Eds.): AH 2006, LNCS 4018, pp. 457–461, 2006. c Springer-Verlag Berlin Heidelberg 2006 

458

M. Rey-López, A. Fernández-Vilas, and R.P. Díaz-Redondo

so as it not only acts as a means of transmission for the courses, on the contrary, the education offered is specific for this medium, taking into account its restrictions and making the most of its potential. Concerning the restrictions, we have to bear in mind both social and technological ones. As the student has just been a viewer for a long time, he/she will probably have a passive attitude when interacting with TV, that is why we have to make education attractive to activate him/her. On the other hand, the contents shown should be in accordance with the technological constraints of IDTV, such as the low resolution of the screen, the fact of using a simple remote control to interact with the programs or the limited features of a set-top box compared with a computer. Considering these limitations, we will take advantage of the fact that viewers have always conceived TV as a pastime and we will try to offer them education without forgetting entertainment. For this to be possible, we will create learning experiences that combine learning elements and audiovisual ones, i.e. TV programs.This way, we obtain two different types of experiences, those having a TV program as its central axis and the ones whose core is a learning element. Another characteristic of these experiences is their personalization according to user’s preferences and learning background, which is essential in t-learning. In this environment, personalization permits the user to access those contents that are interesting for him/her and prevent him/her from getting lost in the huge amount of contents received, compensating in some extent the typical passivity of the viewer. To compose these experiences, we propose the creation of an Intelligent Tutoring System (ITS), which selects, relates and personalizes audiovisual and learning contents. The design of this ITS constitutes the main goal of the Ph.D. work described in this paper, whose objectives will be presented in Section 2. The process to achieve them is explained in Section 3. Finally, in Section 4 we discuss some related research to this topic.

2 Research Objectives We distinguish two types of experiences that our ITS should be able to create. The first one deals with ‘entertainment that educates’. Its central element is a TV program, which will be complemented with learning elements (Fig. 1a). To refer to these experiences, we have applied the term entercation. The construction of entercation experiences is initiated by the selection of a TV program interesting for the viewer. In this moment, the ITS has to choose the most appropriate learning elements —from those ones it has access— related to the characteristics of the program and perceived level of interest for the user. We have to take into account user’s peculiarities to make effective the learning experience and avoid him/her getting bored. The selected learning objects will be offered to the user at the appropriate moment during the program and he/she could access them from this moment on. In these experiences, the TV program acts as a hook to engage viewers in education. The second type refers to ‘education that entertains’. Its central axis is a learning element (Fig. 1b), which will be complemented with TV programs (or segments of these ones) in order for the experience to be more entertaining and attractive for the student. We have used the term edutainment for these experiences. The ITS will create an edutainment experience from a learning element it considers appropriate for the

A Model for Personalized Learning Through IDTV

TV PROGRAMS SELECTION LE TV TV LE LE LE TV

TV

TV

TV

TV

LE

LE

LE

LE

TV

LE

LE

LE

LE

LE LEARNING CONTENT SELECTION

LE

LE

LE

LE

LE

TV

LE

LE

LE

LE

(c) Adaptive entercation

TV

COMPOSITION SELECTION

LE

459

(a) Entercation

LE LE

ADAPTATION LE LE

A LE

LE A LE

IN RN

TV

ITS

G

TV

T EN EM EL TV

IN RN

TV

G

TV

T EN EM EL TV

(b) Edutainment

(d) Adaptive edutainment

Fig. 1. Creation of t-learning experiences

student, according to his/her learning interests. At the appropriate point, it will add some relevant audiovisual elements (that may be of interest to the student), according to learning content, in order to make the experience more entertaining. In order to go a step further towards the personalization of learning experiences for IDTV —the main goal of this Ph.D. work— we intend to add adaptivity to these experiences, obtaining adaptive entercation and edutainment experiences (Figs. 1c and 1d). In these experiences, we introduce adaptive elements, which are adapted in order for the student to achieve its objectives in an appropriate way according to his/her characteristics [6].

3 Research Methodology To achieve the project objectives, an environment based on widely accepted standards is advisable in order for reusability of components and interoperability between systems to be possible. The ITS will work within the technological context defined by the MHP (Multimedia Home Platform) standard [7], which is consolidating worldwide as one of the technical solutions that will shape the future of IDTV. It defines an open interoperable solution that normalizes the characteristics of the set-top boxes and the applications they can execute. On the other hand, the learning elements used by this ITS will comply with the ADL SCORM (Sharable Content Object Reference Model) standard [8], which brings together the works of several normalization initiatives into a consistent body of specifications that is achieving global acceptance. In Fig. 1, we can see the different stages needed to create the proposed learning experiences. The first stage refers to the selection of those contents that are appropriate for the user’s preferences stored in his/her profile. In t-learning, this profile is doublesided since it takes into account the user’s characteristics as a viewer and a student. The system in charge of selecting audiovisual content is a recommender for TV programs. For our work, we will use AVATAR [9], a recommender based on semantic reasoning designed by our research group. For our Ph.D., AVATAR’s most relevant elements are a viewer profile and an ontology based on the TV-Anytime metadata [10], which permits classifying TV programs.

460

M. Rey-López, A. Fernández-Vilas, and R.P. Díaz-Redondo

Regarding learning content, its selection is one of the tasks that our ITS has to perform. For this to be possible, we have to define a user model that reflects his/her preferences and background as a student. To relate this profile with the appropriate learning elements, we have to define an ontology based on SCORM, where the ITS creates the instances of all the elements it can access, thus allowing semantic reasoning. When finalizing the selection process, those elements that are not appropriate for the student have already been discarded. The next phase consists of creating learning experiences from TV programs and learning elements. First, adaptation should be performed for adaptive learning elements, selecting the most appropriate way —among those possible— for the user to achieve the intended objective. Since SCORM does not currently permit adaptivity, we are working on an extension to this standard to achieve learning contents adaptation. This extension should include some structures that provide adaptation rules. This rules allow the ITS to decide which organizations and elements are more appropriate for the characteristics of the user. The last stage is the composition of learning experiences. In this phase, the contents are linked to be shown to the user, semantically relating the instances of learning and audiovisual elements using the aforementioned ontologies, by means of a gateway ontology that contains the concepts of the subject domain.

4 Related Work and Discussion In terms of related work, there are several relevant research fields, including Adaptive Hypermedia, User Modelling and Semantic Reasoning. Adaptive Hypermedia (AH) is one of the most promising areas to offer personalization on the e-learning field. It tries to overcome the problem of having users with different goals and knowledge by using the information represented in the user model to adapt the contents [6]. This is the objective we want to achieve by defining adaptable learning elements appropriate for IDTV. As stated in [11] the techniques used in AH can be extended to audiovisual contents in the field of interactive television. To achieve this adaptation, we are working on extending the SCORM standard with adaptation rules. In this sense, the proposals exposed in [12] and [13] are close to ours since they try to offer adaptivity by including dedicated adaptation-specific constructs in the course definition. However they do not offer different possibilities for users with different needs. With reference to User Modelling, we can find several proposals. One method widely used to represent this one is the overlay model, where the learner knowledge is represented as a subset of the expert knowledge [14]. Another popular method is classifying users into categories and making predictions about them based on a stereotype associated with each category [15]. Regarding the viewer profile, the user model defined for AVATAR stores those branches of the TV ontology that contain the programs the user has already watched [16]. We intend to design our user profile by taking these proposals into account as well as the most relevant standards concerning learner information: IMS LIP (Learner Information Package) [17] and IEEE PAPI (Personal and Private Information) [18]. Regarding Semantic Reasoning, for the selection of learning elements, we take as a starting point the work developed by our group [9], with experience in collaborative

A Model for Personalized Learning Through IDTV

461

filtering and semantic inference that perfectly apply to our needs. To establish relationships between TV programs and learning content, we need a gateway ontology, e.g. SUMO (Suggested Upper Merged Ontology) [19]. To conclude, in this paper we look for a new conception of learning experiences for IDTV combining audiovisual and learning contents, essential requisites for t-learning, taking advantage of its potential instead of using it as a simple means of transmission for the courses. To construct these experiences we will design an ITS to select, relate and personalize the contents, which will be developed using an agent-based architecture. Up to now, we have already designed the SCORM ontology, we are putting the final touches to the SCORM extension and we have developed an authoring tool to create adaptive elements. In the future, we should work on relating learning contents and TV programs to produce entercation and edutainment experiences.

References 1. Bates, P.J.: A Study into TV-based Interactive Learning to the Home. http://www.pjb.co.uk/t-learning (2003) 2. Damásio, M.J., Quico, C.: T-Learning and Interactive Television Edutainment: the Portuguese Case Study. In: Proc. of ED-MEDIA 2004. (2004) 3. Zhao, L.: Interactive Television in Distance Education: Benefits and Compromises. In: Proc. of ISTAS’02. (2002) 255–261 4. Aarreniemi-Jokipelto, P.: Experiences with an Interactive Learning Environment in Digital TV. In: Proc. of ICALT’04. (2004) 5. Pazos-Arias, J.J., et al.: ATLAS: A framework to provide multiuser and distributed t-learning services over MHP. Software: Practice and Experience. (In press) 6. Brusilovsky, P.: Methods and techniques of adaptive hypermedia. User Modeling and User Adapted Interaction 6(2-3) (1996) 87–129 7. DVB Consortium: Multimedia Home Platform 1.2.1. ETSI TS 102 812 V1.2.1 (2003) 8. ADL: Sharable Content Object Reference Model. http://www.adlnet.org (2004) 9. Blanco-Fernández, Y., et al.: AVATAR: An advanced Multi-Agent Recommender System of Personalized TV Contents by Semantic Reasoning. In: Proc. of WISE 2004. (2004) 10. The TV-Anytime Forum: Broadcast and On-line Services: Search, select and rightful use of content on personal storage systems. ETSI TS 102 822 (2004) 11. Masthoff, J., Pemberton, L.: Adaptive hypermedia for personalized TV. In: Adaptable and Adaptive Hypermedia Systems. IDEA group publishing (2005) 246–263 12. Conlan, O.: The Multi-Model, Metadata Driven Approach to Personalised eLearning Services. PhD thesis, Trinity College, Dublin (2005) 13. Mödritscher, F., et al.: Enhancement of SCORM to support adaptive E-Learning within the Scope of the Research Project AdeLE. In: Proc. of ELEARN 2004, USA (2004) 14. Weber, G., Brusilovsky, P.: ELM-ART: An Adaptive Versatile System for Web-based Intruction. International Journal of Artificial Intelligence in Education 1 (2001) 351–384 15. Kobsa, A., et al.: Personalised hypermedia presentation techniques for improving online customer relationships. The Knowledge Engineering Review 16(2) (2001) 111–155 16. Blanco-Fernández, Y., et al.: AVATAR: Modeling Users by Dynamic Ontologies in a TV Recommender System based on Semantic Reasoning. In: Proc. of EuroITV-05. (2005) 17. IMS: Learner Information Package (LIP). http://imsproject.org (2001) 18. IEEE LTSC: Private and Public Information. http://edutool.com/papi/ (2001) 19. Niles, I., Pease, A.: Towards a Standard Upper Ontology. In Welty, C., Smith, B., eds.: Proc. of FOIS-2001, USA (2001)