User tagging of video content provides many possibilities for indexing and personalization. To exploit these possibilities, users must be willing to tag the video content they watch. In this paper we present the first results of our ongoing research, by constructing an overview of user motives to tag video content. We present the results of a study in which we elicited possible user motives to tag movies on the internet. The identified motives include the categories ‘indexing’, ‘socializing’ and ‘communicating’. Finally, user barriers to tag video content are discussed.
In order to navigate huge document collections efficiently, tagged hierarchical structures can be used. For users, it is important to correctly interpret tag combinations. In this paper, we propose the usage of tag groups for addressing this issue and an algorithm that is able to extract these automatically for text documents. The approach is based on the diversity of content in a document collection. For evaluation, we use methods from ontology evaluation and showed the validity of our approach on a benchmark dataset.
This paper studies the properties of a helpful and trustworthy explanation in a movie recommender system. It discusses the results of an experiment based on a natural language explanation prototype. The explanations were varied according to three factors: degree of personalization, polarity and expression of unknown movie features. Personalized explanations were not found to be significantly more Effective than non-personalized, or baseline explanations. Rather, explanations in all three conditions performed surprisingly well. We also found that participants evaluated the explanations themselves most highly in the personalized, feature-based condition.
iFanzy is a personalized TV guide application aiming at offering users television content in a personalized and context-sensitive way. It consists of a client-server system with multiple clients and devices such that the user can ubiquitously use TV set-top box, mobile phone and Web-based applications to select and receive personalised TV content. TV content and background data from various heterogeneous sources is integrated to provide a transparent knowledge structure, which allows the user to navigate and browse the vast content sets nowadays available. Semantic Web techniques are applied for enriching and aligning Web data and (live) broadcast content. The resulting RDF/OWL knowledge structure is the basis for iFanzy's main functionality, like semantic search of the broadcast content and execution of context-sensitive recommendations.
This paper presents an approach to exploit widely used tag annotations to address two important issues in user-adaptive systems: the cold-start problem and the integration of distributed user models. The paper provides an example of re-use of user interaction data (tags) generated by one application into another one in similar domains for providing cross-system recommendations.
We introduce the Second Life Location Recommender System (SLLoRS). This system lets users rate and tag locations within the 3D environment Second Life in order to provide personalized recommendations on a collaborative basis. We demonstrate the system as an in-world application and explore some of the general challenges of applying recommendation systems to 3D online environments, like the implementation of data-intensive applications facing restricted computational resources and the segmentation of recommendations in a continuous input space.
The design of Adaptive Hypermedia is a difficult task which can be made easier if generic systems and AH creators' models are reused. We address this design problem in the setting of the GLAM platform only made up of generic components. In this paper, we assume the GLAM platform is used to create a specific adaptive hypermedia. We present a pattern and a rule-based approach helping a AH creator in reusing its user and domain models and instances in order to make them taken into account. This semi-automatic approach takes the creator's models as specialisations of GLAM generic models and requires the creator to express a minimum set of mappings between his models and the generic ones. The process results in a merged model consisting of the generic and the corresponding specific model, being fully compliant with the GLAM adaptation model. A plug-in and experimentations in the e-learning domain have been partially designed.
Nowadays, the idea of personalization is regarded as crucial in many areas. This requires quick and robust approaches for developing reliable user models. The next generation user models will be distributed (segments of the user model will be stored by different applications) and interoperable (systems will be able to exchange and use user model fractions to enrich user experiences). We propose a new approach to deal with one of the key challenges of interoperable distributed user models - semantic heterogeneity. The paper presents algorithms to automate the user model exchange across applications based on evidential reasoning and advances in the Semantic Web.
Web browsing is a complex activity and in general, users are not guided during browsing. The aim of this research is to support the browsing of users using semantic and adaptive hyperlinks using Semantic Web technologies and personalization methods. In this paper, we propose a novel Semantic Web browser (SemWeB), which uses a behavior-based and an ontology-driven user modeling architecture. In our approach, semantic links and adaptive hypermedia can be achieved on different websites. In addition, user profiles can be easily updated with semantic metadata coming from the Semantic Web browser.
In this paper, we look at initial results of data mining students' help-seeking behaviour in two ITSs: SQL-Tutor and EER-Tutor. We categorised help given by these tutors into high-level (HLH) and low-level help (LLH), depending on the amount of help given. Each student was grouped into one of ten groups based on the frequency with which they used HLH. Learning curves were then plotted for each group. We asked the question, "Does a student's help-seeking behaviour (especially the frequency with which they use HLH) affect learning?"We noticed similarities between results for both tutors. Students who were very frequent users of HLH showed the lowest learning, both in learning rates and depth of knowledge. Students who were low to medium users of HLH showed the highest learning rates. Least frequent users of HLH had lower learning rates but showed higher depth of knowledge and a lower initial error rate, suggesting higher initial expertise. These initial results could suggest favouring pedagogical strategies that provide low to medium HLH to certain students.
In this paper we present a rule-based personalization framework for encapsulating and combining personalization algorithms known from adaptive hypermedia and recommender systems. We show how this personalization framework can be integrated into existing systems by example of the educational online board Comtella-D, which exploits the framework for recommending relevant discussions to the users. In our evaluations we compare different recommender strategies, investigate usage behavior over time, and show that a small amount of user data is sufficient to generate precise recommendations.
The CAWE framework supports the development of context-aware, Service Oriented applications which integrate heterogeneous services and customize the cooperation among multiple users. We present the techniques adopted in the framework to manage a context-sensitive interaction with the users.
This paper presents the methods used in a TV Recommender System that helps users in the difficult task of finding an interesting TV program from among the hundreds of channels that we can find nowadays on TV. Our aim is to cover not only user preferences but also user restrictions while watching TV. The recommendations use a hybrid method, combining content based and folksonomy (collaborative and social recommendations). We also present interesting initial results of some experiments that try to show the accuracy of the users recommendations.
Today, Portals provide users with a central point of access to companywide information. Initially they focused on presenting the most valuable and widely used information to users for efficient information access. But the amount of information accessible quickly grew and finding the right information can hence become a tedious task. We will demonstrate a solution for adapting the Portal's structure, especially its navigation and page structures. We allow for advanced adaptations that each user can perform manually as well as for automated adaptations based on user- and context models reflecting users' interests and preferences. Our main concepts have been embedded and evaluated within IBM's WebSphere Portal.
Open learner models (OLM) are learner models that are accessible to the learner they represent. Many examples now exist, often with the aim of prompting learner reflection on their knowledge. In language learning, this relates to research on noticing and awareness-raising. We here introduce an open learner model to investigate the potential of OLMs to facilitate noticing. Results suggest that an OLM could be a useful way of helping students to notice language features, with all students noticing some of the features tested, a result that was maintained in a delayed post-test one week after the experimental session.
The vast amounts of information presented in museums can be overwhelming to a visitor, whose receptivity and time are typically limited. Hence, s/he might have difficulties selecting interesting exhibits to view within the available time. Mobile, context-aware guides offer the opportunity to improve a visitor's experience by recommending exhibits of interest, and personalising the delivered content. The first step in this recommendation process is the accurate prediction of a visitor's activities and preferences. In this paper, we present two adaptive collaborative models for predicting a visitor's next locations in a museum, and an ensemble model that combines their predictions. Our experimental results from a study using a small dataset of museum visits are encouraging, with the ensemble model yielding the best performance overall.
The motivation behind many Information Retrieval systems is to identify and present relevant information to people given their current goals and needs. Learning about user preferences and access patterns recent technologies make it possible to model user information needs and adapt services to meet these needs. In previous work we have presented ASSIST, a general-purpose platform which incorporates various types of social support into existing information access systems and reported on our deployment experience in a highly goal driven environment (ACM Digital Library). In this work we present our experiences in applying ASSIST to a domain where goals are less focused and where casual exploration is more dominant; YouTube. We present a general study of YouTube access patterns and detail how the ASSIST architecture affected the access patterns of users in this domain.
Misconceptions have been identified in many subjects. However, there has been less investigation into students' interest in their misconceptions. This paper presents two independent open learner models used alongside seven university courses to highlight the state of their knowledge to the learner as a starting point for their independent study. Many students used the environments; many had misconceptions identified at some point during their learning; and most of those with misconceptions viewed the statements of their misconceptions. Students were able to use the independent open learner models in a variety of ways to suit their interaction preferences, at different levels of study.
This paper presents an approach to achieve User Model (UM) interoperability exploiting Web Service technologies for syntactic interoperability, and Semantic Web languages for semantic interoperability, together with negotiation techniques based on dialogue. We propose a SOA-based framework where a central UDDI registry, enhanced with UM specific capabilities, is used to support and promote the cooperation between UM-based applications.
In a recent study, we discovered a new effect of adaptive navigation support in the context of E-learning: the ability to motivate students to work more with non-mandatory educational content. The results presented in this paper extend the limits of our earlier findings. We describe the implementation of adaptive navigation support for the SQL domain, and report the results of the classroom evaluation of our approach. Among other issues, we investigate whether the use in parallel of two different types of navigation support could change the nature or the magnitude of the previously observed effect. Our study confirms the motivational value of navigation support in the new domain. We observe the increase of this effect after adding the concept-based navigation layer to the existing topic-based adaptive navigation service. The results of the navigational pattern analysis allow us to determine the major source of this increase.
