HCI Bibliography Home | HCI Conferences | UMAP Archive | Detailed Records | RefWorks | EndNote | Hide Abstracts
UMAP Tables of Contents: 00010203040506070809101112131415

Proceedings of AH 2002 Adaptive Hypermedia and Adaptive Web-based Systems 2002-05-29

Fullname:Proceedings of the 2rd International Conference on Adaptive Hypermedia and Adaptive Web-Based Systems
Editors:Paul De Bra; Peter Brusilovsky; Ricardo Conejo
Location:Málaga, Spain
Dates:2002-May-29 to 2002-May-31
Series:Lecture Notes in Computer Science, 2002, Volume 2347
Standard No:ISBN: 978-3-540-43737-6 (Print) 978-3-540-47952-9 (Online); hcibib: UMAP02
Links:Online Proceedings
Summary:Hypermedia as a means for accessing information was first proposed by Vannevar Bush, in 1945, in his memorable article "As We May Think". The first research that resembles what we consider as hypermedia (i.e. at least using computers with textual or graphical displays) was started much later, by great minds including Doug Engelbart (known for many inventions including the mouse) and Ted Nelson (known for inventing Xanadu, an architecture similar but superior to that of the World Wide Web). After 20 years of hypermedia research the Web started and really changed hypermedia from a small research community and mostly small stand-alone applications to a world-wide network of interconnected information sources. Hypermedia applications serving a large audience suffer greatly from the traditional "one-size-fits-all" approach of static websites. Starting in the 1990s, and essentially in parallel with the development of the Web, many research teams began to investigate ways of modeling features of the users of hypermedia systems. This has led to a number of interesting adaptation techniques and adaptive hypermedia systems. Nowadays adaptive hypermedia and the Web have really come together. Virtually all adaptive systems use the Web for both the (browser) front end and the server back end.
    Following a number of successful workshops on adaptive hypermedia, the first conference on Adaptive Hypermedia and Adaptive Web-Based Systems (AH 2000) was held in Trento, Italy, in August 2000. (See Springer Lecture Notes in Computer Science, vol. 1892.) The research community has continued to grow. The Second International Conference on Adaptive Hypermedia and Adaptive Web-Based Systems (AH 2002) attracted twice as many full paper submissions, and also a large number of short papers describing ongoing research.
    AH 2002 continued the tradition of earlier workshops, and the AH 2000 conference, in being a meeting place for researchers with very different backgrounds such as hypertext, user modeling, machine learning, natural language generation, information retrieval, intelligent tutoring systems, cognitive science, and Web-based education. In 2001 the community gathered at the User Modeling Conference in Sonthofen, Germany and at the ACM Hypertext Conference in Aarhus, Denmark for a two-session workshop on adaptive hypermedia, reported on in LNCS vol. 2266. Alot of research that was ongoing in 2001 can now be found in the full papers of the AH 2002 conference proceedings. A lot of new research is presented in short papers and posters.
  1. Invited Papers
  2. Full Papers
  3. Short Papers
  4. Posters
  5. Doctoral Consortium

Invited Papers

Is Personalization All About Technology? BIBAFull-Text 1
  Jack Aaronson
A lot of research and development has pushed the limits of personalization far beyond where they were just a few years ago. Yet the major goals of personalization continue to elude us: serving customers what they want, when they want it. Is all this research being developed with pure science in mind, or do the new technologies actually solve real business needs? In this lecture, we will discuss the business needs around personalization. We will explore which technologies have been effective, and which haven't been. We will discuss the importance of mixing technology with smart design to ensure that personalized features are intuitive and easy-to-use. We will also talk about personalization that doesn't require technology.
Adaptive Linking between Text and Photos Using Common Sense Reasoning BIBAFull-Text 2-11
  Henry Lieberman; Hugo Liu
In a hypermedia authoring task, an author often wants to set up meaningful connections between different media, such as text and photographs. To facilitate this task, it is helpful to have a software agent dynamically adapt the presentation of a media database to the user's authoring activities, and look for opportunities for annotation and retrieval. However, potential connections are often missed because of differences in vocabulary or semantic connections that are "obvious" to people but that might not be explicit.
   ARIA (Annotation and Retrieval Integration Agent) is a software agent that acts an assistant to a user writing e-mail or Web pages. As the user types a story, it does continuous retrieval and ranking on a photo database. It can use descriptions in the story to semi-automatically annotate pictures. To improve the associations beyond simple keyword matching, we use natural language parsing techniques to extract important roles played by text, such as "who, what, where, when". Since many of the photos depict common everyday situations such as weddings or recitals, we use a common sense knowledge base, Open Mind, to fill in semantic gaps that might otherwise prevent successful associations.
Resource-Adaptive Interfaces to Hybrid Navigation Systems BIBAFull-Text 12-13
  Wolfgang Wahlster
We present the hybrid navigation systems REAL [1] and SmartKom [2] that combine various positioning technologies to determine the user's location in outdoor and indoor situations and that adapt the multimodal presentation of spatial information to the limited technical resources of various output devices and to the limited cognitive resources of the user.

Full Papers

Ubiquitous User Assistance in a Tourist Information Server BIBAFull-Text 14-23
  Liliana Ardissono; Anna Goy; Giovanna Petrone; Marino Segnan; Pietro Torasso
This paper discusses the provision of ubiquitous services for desktop and handheld interfaces. The focus of the paper is on the interactive agenda offered by INTRIGUE, a tourist information server which assists the user in the organization of a tour by providing personalized recommendations of tourist attractions and helping the user to schedule her itinerary. The interactive agenda can be accessed both remotely, by interacting with the central server, and locally to the user's device.
Automatic Extraction of Semantically-Meaningful Information from the Web BIBAKFull-Text 24-35
  J. Arjona; R. Corchuelo; A. Ruiz; M. Toro
The semantic Web will bring meaning to the Internet, making it possible for web agents to understand the information it contains. However, current trends seem to suggest that the semantic web is not likely to be adopted in the forthcoming years. In this sense, meaningful information extraction from the web becomes a handicap for web agents. In this article, we present a framework for automatic extraction of semantically-meaningful information from the current web. Separating the extraction process from the business logic of an agent enhances modularity, adaptability, and maintainability. Our approach is novel in that it combines different technologies to extract information, surf the web and automatically adapt to web changes.
Keywords: Web agents; information extraction; wrappers; ontologies
Towards Open Adaptive Hypermedia BIBAKFull-Text 36-46
  Christopher Bailey; Wendy Hall; David E. Millard; Mark J. Weal
Research at the University of Southampton has extended generalised Open Hypermedia (OH) models to include concepts of context and behaviour, both traditionally very important to the Adaptive Hypermedia (AH) domain. In this paper we re-evaluate Brusilovsky's pragmatic taxonomy of AH techniques from a structural perspective. A variety of OH structures are discussed that can be used to implement the techniques found in the taxonomy. By identifying common structures we gain a new perspective on the relationship between different AH techniques.
Keywords: Open Hypermedia; Adaptive Techniques; Hypermedia Structure; FOHM; taxonomy
GAS: Group Adaptive System BIBAFull-Text 47-57
  Maria Barra; Paul Maglio; Alberto Negro; Vittorio Scarano
This paper describes an ongoing research project to design a Group Adaptive System (GAS) for collaborative navigation on the web. Our objective is to provide information that adapts to web users based on automatically determined, dynamic user groups. To do this, our approach takes into account the structure of the web (using the CLEVER algorithm [16]) and user interactions when navigating the web (gathered by our collaborative environment) to offer users in a consistent information space determined by their interests and activities. In the end, our system, GAS, enables users to perform asynchronous collaborative navigation by combining resources discovered by a group of users and suggestions discovered by the CLEVER algorithm to provide recommendations to the group.
TV Scout: Lowering the Entry Barrier to Personalized TV Program Recommendation BIBAFull-Text 58-68
  Patrick Baudisch; Lars Brueckner
In this paper, we present TV Scout, a recommendation system providing users with personalized TV schedules. The TV Scout architecture addresses the "cold-start" problem of information filtering systems, i.e. that filtering systems have to gather information about the user's interests before they can compute personalized recommendations. Traditionally, gathering this information involves upfront user effort, resulting in a substantial entry barrier. TV Scout is designed to avoid this problem by presenting itself to new users not as a filtering system, but as a retrieval system where all user effort leads to an immediate result. While users are dealing with this retrieval functionality, the system continuously and unobtrusively gathers information about the user's interests from implicit feedback and gradually evolves into a filtering system. An analysis of log file data gathered with over 10,000 registered online users shows that over 85% of all first-time users logged in again, suggesting that the described architecture is successful in lowering the entry barrier.
Adaptivity, Adaptability, and Reading Behaviour: Some Results from the Evaluation of a Dynamic Hypertext System BIBAFull-Text 69-78
  Kalina Bontcheva
This paper presents the evaluation of a dynamic hypertext system that generates personalised explanations. The main focus is on some of the results, which showed that future experiments need to control for users' reading behaviour, as well as other factors.
Towards Generic Adaptive Systems: Analysis of a Case Study BIBAFull-Text 79-89
  Licia Calvi; Alexandra Cristea
This paper analyses one of the most well-known general purpose adaptive hypermedia systems, AHA!, and, based on this analysis, make some general observations about adaptive hypermedia systems and some improvement suggestions for the AHA! system. We suggest here a concept-based approach to the structuring of adaptive hypermedia systems, as well as an extension of the well-known rule-based overlay method for user-adaptation. This approach is another step towards flexible generic-purpose adaptive hypermedia.
A Methodology for Developing Adaptive Educational-Game Environments BIBAFull-Text 90-99
  Rosa M. Carro; Ana M. Breda; Gladys Castillo; Antonio L. Bajuelos
In this paper we present a methodology for describing adaptive educational-game environments and a model that supports the environment design process. These environments combine the advantages of educational games with those derived from the adaptation. The proposed methodology allows the specification of educational methods that can be used for the game environment generation. The educational goals, the activities that the users can perform, their organization and sequencing, along with the games to be played and the game stories are selected or dynamically generated taking into account the user's features and behaviors.
Multi-model, Metadata Driven Approach to Adaptive Hypermedia Services for Personalized eLearning BIBAFull-Text 100-111
  Owen Conlan; Vincent Wade; Catherine Bruen; Mark Gargan
One of the major obstacles in developing quality eLearning content is the substantial development costs involved and development time required [12]. Educational providers, such as those in the university sector and corporate learning, are under increasing pressure to enhance the pedagogical quality and technical richness of their course offerings while at the same time achieving improved return on investment. One means of enhancing the educational impact of eLearning courses, while still optimizing the return on investment, is to facilitate the personalization and repurposing of learning objects across multiple related courses. However, eLearning courses typically differ strongly in ethos, learning goals and pedagogical approach whilst learners, even within the same course, may have different personal learning goals, motivations, prior knowledge and learning style preferences. This paper proposes an innovative multi-model approach to the dynamic composition and delivery of personalized learning utilizing reusable learning objects. The paper describes an adaptive metadata driven engine that composes, at runtime, tailored educational experiences across a single content base. This paper presents the theoretical models, design and implementation of the adaptive hypermedia educational service. This service is currently being successfully used for the delivery of undergraduate degree courses in Trinity College, Dublin as well as being used as part of a major EU research trial.
Adaptation and Personalization on Board Cars: A Framework and Its Application to Tourist Services BIBAFull-Text 112-121
  L. Console; S. Gioria; I. Lombardi; V. Surano; I. Torre
In this paper we analyse the goals and problems that should be taken into account when designing adaptive/personalized services that must run onboard vehicles. This is, in fact, a very interesting and promising area of application where adaptation and personalization can provide unique advantages. We then introduce a framework and a multi-agent architecture for on-board services supporting different forms of user and context modelling and different forms of adaptation and personalization. Finally, to support our claims and framework, we discuss a specific prototype system for on-board tourist services.
Adaptive Authoring of Adaptive Educational Hypermedia BIBAFull-Text 122-132
  Alexandra Cristea; Lora Aroyo
In this paper we propose a set of development guidelines for an adaptive authoring environment of adaptive educational hypermedia. This set consists of relevant and necessary functionalities and architectural features of authoring systems (AS) for adaptive teaching and/or learning environments (LE). We extracted the core functionalities by analysing MyEnglishTeacher (MyET) and AIMS -- two independently designed and built AS for adaptable/adaptive LE. The extended core led us to a concept-based layered approach with respect to concept- and attribute-level adaptation, as well as lesson and presentation adaptation. We believe that providing adaptive authoring support for adaptive hypermedia will have a strong impact on the authors' motivation and efficiency in performing their tasks and consequently will increase the popularity of adaptive hypermedia.
Hypermedia Presentation Adaptation on the Semantic Web BIBAFull-Text 133-142
  Flavius Frasincar; Geert-Jan Houben
Web Information Systems (WIS) present up-to-date information on the Web based on data coming from heterogeneous sources. In previous work the Hera methodology was developed to support the design of a WIS. In this paper we target the design of an intelligent WIS. For this reason the Hera methodology is extended with two kinds of hypermedia presentation adaptation: adaptability based on a profile storing device capabilities and user preferences, and adaptivity based on a user model storing the user browsing history. While adaptability is considered to be static, i.e. the presentation is fixed before the browsing starts, adaptivity is dynamic, i.e. the presentation changes while the user is browsing it. The models used in Hera and their adaptation aspects are specified in RDF(S), a flexible Web metadata language designed to support the Semantic Web.
User Data Management and Usage Model Acquisition in an Adaptive Educational Collaborative Environment BIBAFull-Text 143-152
  Elena Gaudioso; Jesus G. Boticario
In this paper we describe the basis of aLF (active learning framework) an environment for web-based distance learning on the web and how it can be extended with a user modelling subsystem aimed at providing an adaptive response which can meet users' knowledge state, preferences and goals. We concentrate on user data collection and its further processing to create a user model. Furthermore, as regards the user modelling task, we claim that a combination of different machine learning algorithms should perform better than individual methods. To support our approach we will provide some preliminary experimental results.
Personalizing Assessment in Adaptive Educational Hypermedia Systems BIBAFull-Text 153-163
  Evangelia Gouli; Kyparisia Papanikolaou; Maria Grigoriadou
In this paper, we present a comprehensive framework for assessment, developed through the web-based module named PASS-Personalized ASSessment, which can be integrated in an Adaptive Educational Hypermedia System to provide personalized assessment. PASS estimates learner's performance through multiple assessment options -- pre-test, self-assessment and summative assessment -- tailored to learner's responses. The adaptive functionality of PASS, which is mainly based on the adaptive testing and the adaptive questions techniques, is described. The first results from the formative evaluation of PASS are encouraging, concerning the total number of questions posed to estimate learner's knowledge level, which is usually less than the maximum needed and the accuracy of the outcome results compared to the estimations of the expert-tutor.
Visual Based Content Understanding towards Web Adaptation BIBAFull-Text 164-173
  Xiao-Dong Gu; Jinlin Chen; Wei-Ying Ma; Guo-Liang Chen
Web content structure is proposed to facilitate automatic web page adaptation in this paper. By identifying the logic relationship of web content based on layout information, web content structure effectively represents authors' presentation intention. An automatic top-down, tag-tree independent approach to detect web content structure is presented. It simulates how a user understands web layout structure based on his vision. Comparing to other content analysis techniques, our approach is independent to physical realization and works well even when the physical structure is far different from layout structure. Besides, our approach is an O(n)-time process which is much more efficient comparing to other approaches with O(n²)-time complexity. Furthermore, our approach is tag tree independent, which means it can be applied to web contents of arbitrary physical realization formats. Experiments show satisfactory results.
Knowledge Modeling for Open Adaptive Hypermedia BIBAFull-Text 174-183
  Nicola Henze; Wolfgang Nejdl
This paper proposes a knowledge modeling approach for adaptive, open corpus hypermedia systems. Our approach towards adaptive, open corpus hypermedia is based on interpreting standard metadata of learning objects. For each corpus of documents integrated into the open adaptive hypermedia system (OAHS) we are calculating subgraphs of the ontology for estimating the user's knowledge with respect to this corpus. This enables the OAHS to understand the knowledge contained in learning materials, to make estimations about an individual user's knowledge state and to learn the prerequisite knowledge required for learning objects from given structures in the materials.
Adaptive Navigation for Learners in Hypermedia Is Scaffolded Navigation BIBAFull-Text 184-192
  Roland Hübscher; Sadhana Puntambekar
Adaptive navigation support can be of great help in large hypermedia systems supporting learners as well as users searching for specific information. A wide variety of adaptive mechanisms have been implemented in existing adaptive hypermedia systems that provide better and better suggestions to the user what hyperlinks to follow. We suggest that adaptive navigation support should scaffold a learner in an educational hypermedia system to select the appropriate links. We show that this implies that selecting a link is an educationally relevant activity that should not always be reduced to a trivial task by powerful adaptive mechanisms. It follows that learners require sometimes different kinds of adaptive navigation support than users looking for information. Finally, we will suggest how to extend current mechanisms to provide scaffolded navigation support to learners.
Pros and Cons of Controllability: An Empirical Study BIBAFull-Text 193-202
  Anthony Jameson; Eric Schwarzkopf
A key usability issue with systems that adapt to their users is controllability: the ability of the user to determine the nature and timing of the adaptation. This paper presents an empirical study of the trade-offs involved in an attempt to ensure a suitable degree of controllability. Within an adaptive hotlist for a conference web site, two mechanisms for providing users with recommendations of conference events were compared: automatic vs. controlled updating of recommendations. In an experimental setting, each of 18 users worked with both variants of the adaptive hotlist, as well as with a nonadaptive variant. The users differed markedly in their responses to automatic vs. controlled updating. A number of reasons for these differences could be found in the objective and subjective data yielded by the study. The study illustrates how preferences for different forms of user control can be influenced by factors ranging from stable individual differences to unpredictable features of a situation. General implications for the design of controllable adaptive systems are discussed.
Personis: A Server for User Models BIBAKFull-Text 203-212
  Judy Kay; Bob Kummerfeld; Piers Lauder
A core element of an adaptive hypertext systems is the user model. This paper describes Personis, a user model server. We describe the architecture, design and implementation. We also describe the way that it is intended to operate in conjunction with the rest of an adaptive hypertext system.
   A distinctive aspect of the Personis user model server follows from our concern for making adaptive systems scrutable: these enable users to see the details of the information held about them, the processes used to gather it and the way that it is used to personalise an adaptive hypertext. We describe how the architecture supports this.
   The paper describes our evaluations of the current server. These indicate that the approach and implementation provide a workable server for small to medium sized user collections of information needed to adapt the hypertext.
Keywords: Server for user profile/model; Security and Privacy of User Models; User Modelling; Personalisation; User Control
The Munich Reference Model for Adaptive Hypermedia Applications BIBAKFull-Text 213-222
  Nora Koch; Martin Wirsing
Although adaptive applications are increasing in popularity, there are only a few approaches that focus on their generalization or the specification of a reference model. Trying to fill this gap, this paper presents a reference model for adaptive hypermedia applications, similar to AHAM. The main novelty of our approach is an object-oriented specification written in UML (Unified Modeling Language) which integrates both an intuitive visual representation and a formal unambiguous specification in OCL (Object Constraint Language). Our reference model is defined as an extension of the Dexter Hypertext Reference Model including user modeling aspects and rule-based adaptation mechanisms.
Keywords: Adaptive Hypermedia; Reference Model; Visual Modeling; UML; Formal Specification; Constraint Language; OCL
Tracking Changing User Interests through Prior-Learning of Context BIBAFull-Text 223-232
  Ivan Koychev
The paper presents an algorithm for learning drifting and recurring user interests. The algorithm uses a prior-learning level to find out the current context. After that, searches into past observations for episodes that are relevant to the current context, 'remembers' them and 'forgets' the irrelevant ones. Finally, the algorithm learns only from the selected relevant examples. The experiments conducted with a data set about calendar scheduling recommendations show that the presented algorithm improves significantly the predictive accuracy.
Prediction of Navigation Profiles in a Distributed Internet Environment through Learning of Graph Distributions BIBAFull-Text 233-241
  Dirk Kukulenz
Collaborative filtering techniques in the Internet are a means to make predictions about the behaviour of a certain user based on the observation of former users. Frequently in literature the information that is made use of is contained in the access-log files of Internet servers storing requested data objects. However with additional effort on the server side it is possible to register, from which to which data object a client actually navigates. In this article the profile of a user in a distributed Internet environment will be modeled by the set of his navigation decisions between data objects. Such a set can be regarded as a graph with the nodes beeing the requested data objects and the edges being the decisions. A method is presented to learn the distribution of such graphs based on distance functions between graphs and the application of clustering techniques. The estimated distribution will make it possible to predict future navigation decisions of new users. Results with randomly generated graphs show properties of the new algorithm.
Educo -- A Collaborative Learning Environment Based on Social Navigation BIBAFull-Text 242-252
  Jaakko Kurhila; Miikka Miettinen; Petri Nokelainen; Henry Tirri
Web-based learning is primarily a lonesome activity, even when it involves working in groups. This is due to the fact that the majority of web-based learning relies on asynchronous forms of interacting with other people. In most of the cases, the chat discussion is the only form of synchronous interaction that adds to the feeling that there are other people present in the environment. EDUCO is a system that tries to bring in the sense of other users in a collaborative learning environment by making the other users and their the navigation visible to everyone else in the environment in real-time. The paper describes EDUCO and presents the first empirical evaluation as EDUCO was used in a university course.
GOOSE: A Goal-Oriented Search Engine with Commonsense BIBAFull-Text 253-263
  Hugo Liu; Henry Lieberman; Ted Selker
A novice search engine user may find searching the web for information difficult and frustrating because she may naturally express search goals rather than the topic keywords search engines need. In this paper, we present GOOSE (goal-oriented search engine), an adaptive search engine interface that uses natural language processing to parse a user's search goal, and uses "common sense" reasoning to translate this goal into an effective query. For a source of common sense knowledge, we use Open Mind, a knowledge base of approximately 400,000 simple facts such as "If a pet is sick, take it to the veterinarian" garnered from a Web-wide network of contributors. While we cannot be assured of the robustness of the common sense inference, in a substantial number of cases, GOOSE is more likely to satisfy the user's original search goals than simple keywords or conventional query expansion.
On Adaptability of Web Sites for Visually Handicapped People BIBAFull-Text 264-273
  Mercedes Macías; Julia González; Fernando Sánchez
Currently, the great majority of content published on the Internet is inaccessible to visually impaired users. Although designers have guidelines that guarantee the accessibility of pages constructed as well as software tools to facilitate this task, it is necessary to consider the user's perspective too, allowing him/her to participate in the restructuring or presentation process of contents. There are few software tools which are able to do this. In this paper we present KAI (Accessibility Kit for the Internet) that considers both the user and the designer. It classifies the different components of a published Web page and presents them to the user according to his/her needs. KAI is based on a new language, BML (Blind Markup Language) that helps authors to develop better structured pages. It provides two levels of independence: original Web code and user browser or navigation platform. KAI includes a mixed audio/touch browser (WebTouch) that enables selective reading of contents. The proposed accessibility kit uses several accessibility metrics to ensure that those pages transformed by KAI are more accessible than the original ones. In this paper we give an overview of the overall system.
A Framework for Filtering and Packaging Hypermedia Documents BIBAFull-Text 274-283
  Lucimar C. Martins; Tatiana A. S. Coelho; Simone D. J. Barbosa; Marco A. Casanova
Despite great effort in attempting to develop systems that personalize both content and presentation, there are still some important challenges related to information filtering, packaging and formatting that adapt to user's goals, interests and presentation preferences. This paper addresses these issues by proposing a three-level framework that achieves a high degree of separation of concerns. The framework dissociates the packaging process from the filtering and formatting processes, and thus facilitates the implementation, user testing and fine-tuning of the system representations and algorithms.
Adaptation in an Evolutionary Hypermedia System: Using Semantic and Petri Nets BIBAKFull-Text 284-295
  Nuria Medina-Medina; Lina García-Cabrera; Ma José Rodríguez-Fortiz; José Parets-Llorca
In this paper a classification of adaptive hypermedia systems is presented. Advantages and disadvantages of these systems are also discussed. As a consequence, the need of evolution is argued. An adaptive and evolving hypermedia system is presented and outlined, where the user model is formalized by means of Petri Nets. Finally, an example shows how in our proposal the user model is initialized and updated during the browsing and how adaptation is carried out over the conceptual structure of navigation.
Keywords: Hypermedia System; Adaptation; Petri Net; Evolution; SEM-HP model
Evaluating the Effects of Open Student Models on Learning BIBAFull-Text 296-305
  Antonija Mitrovic; Brent Martin
In previous work [10], we reported on an experiment performed in the context of SQL-Tutor, in which we analysed students' self-assessment skills. This preliminary study revealed that more able students were better in assessing their knowledge. Here we report on a new study performed on the same system. This time, we analysed the effect of an open student model on students' learning and self-assessment skills. Although we have not seen any significant difference in the post-test scores of the control and the experimental group, the less able students from the experimental group have scored significantly higher than the less able students from the control group. The more able students who had access to their models abandoned significantly less problems the control group. These are encouraging results for a very simple open model used in the study, and we believe that a more elaborate model would be more effective.
Ephemeral and Persistent Personalization in Adaptive Information Access to Scholarly Publications on the Web BIBAFull-Text 306-316
  Stefano Mizzaro; Carlo Tasso
We show how personalization techniques can be exploited to implement more adaptive and effective information access systems in electronic publishing. We distinguish persistent (or long term) and ephemeral (or short term) personalization, and we describe how both of them can be profitably applied in information filtering and retrieval systems used, via a specialized Web portal, by physicists in their daily job. By means of several experimental results, we demonstrate that persistent personalization is needed and useful for information filtering systems, and ephemeral personalization leads to more effective and usable information retrieval systems.
Fuzzy Linguistic Summaries in Rule-Based Adaptive Hypermedia Systems BIBAFull-Text 317-327
  Miguel-Ángel Sicilia; Paloma Díaz; Ignacio Aedo; Elena García
Rule-based adaptive hypermedia systems personalize the structure of the hypermedia space using an inference mechanism that operates on a specific knowledge representation about its users. Approximate quantifiers are very frequently used in human language expressions that entail the summarization of a large number of facts. We describe how quantified expressions can be used in adaptation rules to specify common adaptation behaviors, enhancing rule's expressive power for the human expert. Those quantified expressions can be implemented through fuzzy quantification mechanisms operating on fuzzy linguistic labels and relations, and can be integrated as extensions in general-purpose rule-based adaptive hypermedia systems.
The Plight of the Navigator: Solving the Navigation Problem for Wireless Portals BIBAFull-Text 328-337
  Barry Smyth; Paul Cotter
The wireless Internet, as epitomized by the first generation of WAP-enabled phones and portals, has failed to meet user expectations, resulting in limited take-up, and poor revenue growth for mobile operators. A combination of factors has been responsible: unreliable early handsets; poor content; slow connections; and portals that were difficult to use and navigate. Today, the first 3 of these issues have been solved (or are about to be) by improved handsets, high-quality content and high-speed infrastructure such as GPRS. However, portal usability remains a key problem limiting the ease with which users can locate and benefit from wireless content services. In this paper we describe how personalized navigation techniques can greatly enhance the usability of information services in general, and WAP portals in particular, by personalizing the navigation structure of a portal to the learned preferences of individual users, and how this has resulted in increased WAP usage in live user trials.
Towards an Adaptive Web Training Environment Based on Cognitive Style of Learning: An Empirical Approach BIBAFull-Text 338-347
  Maria A. M. Souto; Regina Verdin; Ricardo Wainer; Milton Madeira; Mariusa Warpechowski
The paper presents an investigation about the learner's cognitive profiles in the context of Tapejara Project[1]. In this investigation, we have assumed that the learner's Cognitive Style of Learning (CSL) actually influences his/her cognitive behavior while performing a long distance Web course. The paper describes the empirical procedures used to generate the learners' cognitive profiles and his/her associated learning behavior that must be considered in the adaptive Web training environment. The training target population refers to the Telecommunications Company's employees. The learner's CSL assessment was obtained from the statistical analysis of the Ross Test data [14]. The CSL learning behavior was obtained from the statistical analysis of the navigational log data in an experimental training module designed for this purpose. The statistical results have shown five CSL groups, their style dimensions and the correlation between each CSL group and their navigational behavior in the Web training module.
Automated Personalization of Internet News BIBAFull-Text 348-357
  Aditya V. Sunderam
A systems approach to the automatic and adaptive personalization of Internet news is described. Implemented on the client side as a lightweight, transparent software system, this approach is based on implicit user feedback, thereby preserving privacy while avoiding the constant recustomization needed in explicit schemes. The system consists of two modules: (1) a profiling agent, which unobtrusively monitors news reading patterns to track interest in different topics while acting as a proxy server on the user's computer; (2) and an action agent, which uses the profile information to retrieve, filter, and present news articles. A prototype of the system was implemented and evaluated over a two week period. Precisions (the percentages of relevant articles returned by the system) ranging from 60-95% were observed for profiles representing various combinations of interests. The system also responded very well to simulated changes in user interests, returning rapidly increasing numbers of articles relevant to newly developed interests.

Short Papers

Conceptual Modeling of Personalized Web Applications BIBAFull-Text 358-362
  Silvia Abrahão; Joan Fons; Magalí González; Oscar Pastor
The demand for web applications that take into account the different needs and interests of the users has been increasing. Personalization can involve a wide-range of approaches and techniques to design the end-user experience. This paper discusses the use of conceptual modeling techniques in a software production process for the design of personalized web applications. This process is driven by an Object-Oriented Web-Solutions Modeling approach (OOWS) that properly captures the specific characteristics of web applications. We place special emphasis on the primitives of a navigational model that provides personalization patterns to capture and represent the semantics of this kind of application.
On Evaluating Adaptive Systems for Education BIBAFull-Text 363-367
  Rosa Arruabarrena; Tomás A. Pérez; J. López-Cuadrado; J. Gutiérrez; J. A. Vadillo
In this paper we have gathered some interesting techniques to evaluate educative systems. Our main purpose is to evaluate HEZINET, an adaptive hypermedia system for education available commercially. Therefore, we also include a system evaluation plan to identify which types and techniques of evaluation will be needed in order to accomplish it.
Recommending Internet-Domains Using Trails and Neural Networks BIBAFull-Text 368-371
  Tobias Berka; Wernher Behrendt; Erich Gams; Siegfried Reich
This paper discusses the use of artificial neural networks, trained with patterns extracted from trail data, as recommender systems. Feed-forward Multilayer-Perceptrons trained with the Backpropagation Algorithm were used to assign a rating to pairs of domains, based on the number of people that had traversed between them. The artificial neural network constructed in this project was capable of learning the training set to a great extent, and showed good generalizational capacities.
Learning Grammar with Adaptive Hypertexts: Reading or Searching? BIBAFull-Text 372-375
  Angela Brunstein; Jacqueline Waniek; Anja Naumann; Josef F. Krems
Using adaptive hypertext as a learning tool, the present study addresses the question of the effects of different processing goals, especially the goal of general reading a hypertext as opposed to the goal of searching for specific information, on learning content and skills at English grammar. Twenty students with German as mother tongue processed the present continuous chapter of the Chemnitz InternetGrammar. It has been shown that readers answer a higher amount of questions about details and, more importantly, they answer them in more detail than searchers. Nevertheless, searchers tend to produce more elaborated answers to complex questions than readers. Both groups performed better on skill tests after the session, showing no effect of the performed task. Based on this experimental evidence, the requirements on adaptive learning tools are discussed.
SIGUE: Making Web Courses Adaptive BIBAFull-Text 376-379
  Cristina Carmona; David Bueno; Eduardo Guzman; Ricardo Conejo
Most of the information of the WWW is not adaptive, rather it is dispersed and disorganized. On the other hand, the creation of an adaptive course is a complex task. SIGUE is an author tool that makes possible to build adaptive courses using existing web pages. This means that if there is a lot of information on the web about the same topic the author doesn't have to design the content of a specific course, he can reuse these pages to build his own course, taking the best pages for the concepts he wants to explain. The author can also construct adaptive courses reusing previously non-adaptive ones. SIGUE provides an enhanced interface for the student, controls his interaction, and annotates the visited links in a student model.
An Ontology-Guided Approach to Content Adaptation in LEO: A Learning Environment Organizer BIBAFull-Text 380-383
  John W. Coffey
This paper describes new capabilities of a nonlinear course presentation program entitled LEO: a Learning Environment Organizer. A Knowledge-based Instructional Assistant combines knowledge of the learner's attainment with knowledge of various attributes of the instructional media in order to make recommendations of media that might benefit the learner. This work describes an ontology to represent the basic conceptualization of the learning system, and how that representation is used to match potentially useful instructional content to student profiles. Recommendations of media are based upon multi-parameter similarity measures between media content and capability attainment.
A Scrutable Adaptive Hypertext BIBAFull-Text 384-387
  Marek Czarkowski; Judy Kay
This paper describes Tutor/ADAPT, a working model and framework for scrutable adaptive hypertext. This enables the user to scrutinise the system to determine how the document was adapted. They are also able to control the adaptation by altering the user model at any time in the interaction. The user can see which parts of the user model have been used to control aspects, which appear in the document. The system has three major parts: an authoring tool; the web based interface; and ATML is the language used for adaptive hypertext source documents. This paper describes the user view, especially the scrutability support.
AHA! Meets AHAM BIBAFull-Text 388-391
  Paul De Bra; Ad Aerts; David Smits; Natalia Stash
After many years of hypertext research, the Dexter model was defined [7] to capture the features of most existing hypertext systems in a single, formal reference model. Likewise, the AHAM model [5] (based on Dexter) describes most features that are found in adaptive hypermedia systems (ahs). In the AHA! project funded by the NLnet Foundation we are extending the simple AHA system [4], [6] with the condition-action rules that were introduced in AHAM [8]. This results in a more versatile adaptation engine, with more intuitive and more powerful rules.
Adaptive Real Time Comment Generation for Sail Racing Scenarios BIBAFull-Text 392-396
  Andrea Esuli; Antonio Cisternino; Giuliano Pacini; Maria Simi
We describe a comment generation subsystem developed as part of an expert authoring system for sail racing scenarios. The result of the expert system analysis is sent to a multimedia presentation subsystem composed of a 3D player and a speaker agent. Comments output by the text-to-speech component must be carefully generated so that they are concise, relevant, non repetitive and timely since they must be synchronized with the 3D animation. The comment generation process is adaptive with respect to the time available and to the user profile and preferences.
A Web-Based Selling Agent That Maintains Customer Loyalty through Integrative Negotiation BIBAFull-Text 397-400
  Michael Grimsley; Anthony Meehan
In many transactions, the seller's objective includes promoting customer loyalty in order to increase the likelihood of further transactions. Integrative bargaining styles foster positive relationships between parties. This short paper describes a protocol for a web-based selling agent that adopts an integrative selling style to identify alternative sales contracts that match customer priorities and hence promote customer satisfaction.
Adaptive Content for Device Independent Multi-modal Browser Applications BIBAFull-Text 401-405
  Jennifer Healey; Rafah Hosn; Stéphane H. Maes
Adapting content appropriate to the device and modality of a user's preference becomes more important as users begin to expect universal access to information, whether they are on the phone, on a desktop or using a PDA. This paper outlines the design of a travel application authored using an XForms compliant language and deployed using a DOM-based MVC multi-modal browser. The travel application authored in a language of conversational gestures that can be transcoded into multiple synchronized views for access via a variety of devices.
Behavioral Sequences: A New Log-Coding Scheme for Effective Prediction of Web User Accesses BIBAFull-Text 406-410
  Rushed Kanawati; Maria Malek
Mining web site logs for predicting user actions is a central issue in the field of adaptive web site development. In order to match the dynamic nature of today web sites we propose in this paper a new scheme for coding Web server log data into sessions of behavioral sequences. Following the proposed coding scheme the navigation sessions are coded as a sequence of hypothetical actions that may explain the transition from one page to another. The output of a prediction algorithm will now be an action that can be evaluated in the context of the current navigation in order to find pages that to be visited by the user.
A Fuzzy-Based Approach to User Model Refinement in Adaptive Hypermedia Systems BIBAFull-Text 411-414
  Cveta Martinovska
This paper presents an approach for refinement of the user model in hypermedia applications using fuzzy adaptation rules. The rules register relevant user browsing actions and this way change the strength of relationship between the user model attributes and the concepts of the knowledge domain. The process of fuzzy inference is formalized with fuzzy automaton. Using appropriate authoring tool fuzzy automata provides a method for detecting conflicts and inconsistency in the set of rules, in an automated fashion.
Towards an Authoring Coach for Adaptive Web-Based Instruction BIBAFull-Text 415-418
  Judith Masthoff
A new approach to adaptive web-based instruction is outlined and advocated, which is based on models and principles that allow the course material to organize itself. This is particularly useful when course material originating from multiple authors is to be combined into a coherent whole.
Generation of Personalized Web Courses Using RBAC BIBAFull-Text 419-423
  Susana Montero; Ignacio Aedo; Paloma Díaz
Role-based access control policies model the users domain by means of complex structures where the roles assumed by the users are specialized into more concrete subroles which inherit properties and authorizations from their parents. In this paper, we describe how to combine these principles and adaptation to provide personalized access to the different types of users of a web-based course. We also present Courba, a platform that generates personalized web-based courses using XML to support the definition of access policies.
An Automatic Rating Technique Based on XML Document BIBAFull-Text 424-427
  Hyeonjeong Mun; Sooho Ok; Yongtae Woo
In this paper, we propose an automatic rating technique to collect the interest of users on the contents, such as e-books and electronic catalogs that are composed of XML documents, for developing a personalized recommender system. Our approach focuses on a method to collect implicit rating values for the elements in an XML document accessed by a user, when the content is converted into HTML format. In general, existing implicit rating techniques collect rating values after analyzing access logs in batch mode, however our method can collect the rating values in realtime. As a result of experimentation, we show that the implicit rating values collected by the proposed method are strongly correlated with explicit rating values.
Using Effective Reading Speed to Integrate Adaptivity into Web-Based Learning BIBAFull-Text 428-431
  Muan Hong Ng; Wendy Hall; Pat Maier; Ray Armstrong
It has always been difficult to determine to what extent a user has read a page especially in the area of educational adaptive hypermedia systems. We propose the use of an individual's effective reading speed to estimate how much of a page a user has read during their browsing activity. This method is currently used to apply history-based link annotation in a medical web-based learning application, JointZone. A validation test of this work has shown a positive result in approximating user's reading value when compared to conventional methods.
A Non-invasive Cooperative Student Model BIBAFull-Text 432-435
  Teresa Roselli; Enrica Gentile; Paola Plantamura; Veronica Rossano; Vittorio Saponaro
A crucial issue in building a user-centered Intelligent Educational System is that of building and maintaining user models to enable the system to tailor its behavior to the needs of the given individual. Various student modeling techniques are described in the literature (overlay, perturbation, differential), that represent the student's knowledge in different ways. A common problem affecting these techniques is how to obtain the data needed to build the student model. Open models are one of the possible solutions but these have some drawbacks. In the present article we propose a new open model aiming to solve some of these, by minimizing the invasiveness of the requirement for student involvement in the self-assessment process, together with the sense of frustration such a requirement can arouse in students who are unable to assess themselves and are therefore ignored by the system.
Category-Based Filtering in Recommender Systems for Improved Performance in Dynamic Domains BIBAFull-Text 436-439
  Mikael Sollenborn; Peter Funk
In Recommender systems, collaborative filtering is the most commonly used technique. Although often successful, collaborative filtering encounters the latency problem in domains where items are frequently added, as the users have to review new items before they can be recommended. In this paper a novel approach to reduce the latency problem is proposed, based on category-based filtering and user stereotypes.
Protecting the User from the Data: Security and Privacy Aspects of Public Web Access BIBAFull-Text 440-443
  Adrian Spalka; Hanno Langweg
Observance of laws on the protection of minors is a serious problem at places with Internet access. We examine protection mechanisms in view of the Internet's open and dynamic nature and claim that those based on deterrence and restoration should be preferred to purely preventive ones. We then present a strategy adequate for secondary schools. The technical part relies on adaptable filter and logging components, which check and log requested web-pages. In a feed-back loop a human auditor provides more information to the system, thus increasing the effectiveness. On the other hand, separation of powers among auditors ensures the users' privacy.
Exploiting a Progressive Access Model to Information BIBAFull-Text 444-447
  Marléne Villanova-Oliver; Jérôme Gensel; Hervé Martin
We propose here to stratify the Information Space of a Web-based Information System (WIS) by decomposing it into personalized sub-Information Spaces. This stratification is described through a Progressive Access Model (PAM) written in UML. The PAM gives WIS users, first, access to some minimal and essential information, and then, allows them to navigate through larger and/or smaller personalized Information Spaces. Together with the PAM, we present a specific query language which allows to query a stratification and to format replies according to the different levels of details of the stratified user's Information Space.
Adapting to Prior Knowledge of Learners BIBAFull-Text 448-451
  Stephan Weibelzahl; Gerhard Weber
Prior knowledge is an important factor that influences the interaction with a hypertext and the learning gain. Our authoring system NetCoach provides a way to assess the users' prior knowledge and to adapt the course in different ways. We describe how the adaptation mechanism assesses the user's knowledge with test items, infers the user's current learning state from this information, and finally adapts accordingly. An evaluation study with an adaptive HTML course demonstrates that this kind of adaptation might reduce the completion time, but retains the learning gain.
Monitoring Partial Updates in Web Pages Using Relational Learning BIBAFull-Text 452-455
  Seiji Yamada; Yuki Nakai
This paper describes an automatic monitoring system that constantly checks partial updates in Web pages and notifies the user about them. While one of the most important advantages of the WWW is frequent updates of Web pages, we need to constantly check them out and this task may take much cognitive load. Unfortunately applications to automatically check such updates cannot deal with partial updates like updates in a particular cell of a table in a Web page. Hence we developed an automatic monitoring system that checks such partial updates. The user can give a system regions in which he/she wants to know the updates in a Web page as training examples, and the system is able to learn rules to identify the partial updates by relational learning. We implemented the system and some executed examples were presented.


Adaptation in the Web-Based Logic-ITA BIBAFull-Text 456-461
  David Abraham; Kalina Yacef
The Logic-ITA is an Intelligent Teaching Assistant system for the teaching/learning of propositional logic. The system is tailored to two different types of uses: for students, it is an autonomous Intelligent Tutoring System, whilst for teachers, it includes functionality to set up learning levels, adjust parameters for progressing through these levels, monitor the class' progress and collect data. This paper presents some characteristics of the Web-based version of the Logic-ITA. In particular, the web-based version allows a centralisation on the server of the student models, which contain all the users' individual information related to their learning. This means that all the adaptation features of the Logic Tutor are maintained, whilst providing (1) students with a large degree of flexibility in terms of when and where they choose to use the tool, and (2) teachers with the possibility to monitor accurately the results and statistics.
Collaborative Radio Community BIBAKFull-Text 462-465
  Paolo Avesani; Paolo Massa; Michele Nori; Angelo Susi
Recommender systems have been usually designed to support a single user in a one-to-one relation between a human and a service provider. This paper presents a collaborative radio community where the system delivers a personalization service on the fly, on the basis of the group recommending, promoting a shift from the one-to-one approach to a one-to-group scenario where the goal is assisting people in forming communities.
Keywords: recommender systems; web radio; learning preferences; multicast streaming; virtual community
Improving Interactivity in e-Learning Systems with Multi-agent Architecture BIBAFull-Text 466-471
  Ricardo Azambuja Silveira; Rosa Maria Vicari
Over the last years, many organizations started to use Distance Teaching tools as instruments in employees' qualification programs, creating what we may call E- learning or Virtual Training in Human Resources Development Programs. However, usually these organizations tend to use technological resources already available, and do not shape their technological platform into a pedagogical project. Recent advances in the field of Intelligent Teaching Systems have proposed the use of Artificial Intelligence through architectures based on agents' societies. Teaching systems based on Multi-Agent architectures make possible to support the development of more interactive and adaptable systems. The objective of the paper is to discuss the feasibility of implementing Distributed Intelligent Learning Environment -- DILE based on the Multi-Agents Architecture approach, aiming at the achievement of human resources qualification through Virtual Training. Besides, we present a proposal of an architecture named JADE -- Java Agent Framework for Distance Learning Environments.
Adaptive Hypermedia Made Simple with HTML/XML Style Sheet Selectors BIBAFull-Text 472-475
  François Bry; Michael Kraus
A simple extension is proposed for enhancing HTML and XML with adaptation. It consists in using the path selectors of style sheet languages such as CSS and XSLT for expressing content and navigation adaptation. The needed extensions to a path selector language are minimal, a few additional constructs suffice. The processor of the language can be kept almost unchanged, no new algorithms are needed. Furthermore, it is proposed to use XML for expressing user model data like browsing history, browsing environment (such as device, time, etc.), and application data (such as user performances on exercises).
A Framework for Educational Adaptive Hypermedia Applications BIBAFull-Text 476-479
  Félix Buendía; Paloma Díaz
Hypermedia technology has a large spread in the instructional field, specially in Web-based educational environments. Adaptive hypermedia systems have been focused on such aspects as specifying user models to adapt the educational contents and the navigation structure providing learning paths. The work presented in this paper addresses a new aspect based on adapting the access to the didactic structures, that encapsulate educational contents and learning paths, to support useful educational adaptive applications. It is presented a framework to support the design process which is divided in two main parts: the first one is focused on instructional issues such as the organization of educational contents and the learning tasks to access them; and the second one is based on the translation of these instructional entities into hypermedia components, for which the Labyrinth hypermedia model is used.
METIORE: A Publications Reference for the Adaptive Hypermedia Community BIBAFull-Text 480-483
  David Bueno; Ricardo Conejo; Cristina Carmona; Amos A. David
The Web is one of the most powerful sources of information on any topic. However looking for scientific literature is a difficult task. Prior knowledge of link sites is necessary and if you are lucky they point to conferences proceedings available on-line. In fact the case the user is not able to make queries about the available documents and must check them one by one using general purpose search engines. In this paper we propose our system METIORE as a source of information for the Adaptive Hypermedia community. The idea is to put together all the publications on this research area and provide an adaptive tool to find papers or people working in the field. METIORE is a Personalized Information Retrieval system that keeps a user model based on objectives.
User-Driven Adaptation of e-Documents BIBAFull-Text 484-487
  P. Carrara; D. Fogli; G. Fresta; P. Mussio
This paper proposes a new strategy for designing e-documents adaptable to user's aims and habits. This strategy is based on the creation of a cascade of interaction environments requiring the local adaptation of the e-documents. The strategy is supported by a recently introduced model of Human Computer Interaction, and made effective by the development of BANCO, an environment allowing Web modification by the users at the client side.
Server Independent Personalization for Effective Searching in the World Wide Web BIBAFull-Text 488-491
  Lillian N. Cassel; Ursula Wolz; Robert E. Beck
Effective Web searching involves two components: (1) a search tool must know about and characterize as many Web pages as possible and (2) it must match a user query to the most suitable resource. The essential problem addressed by personalization systems relates to the second component of Web searching. How can the information that a search service holds about resources available on the Web be matched to the diverse needs of the global user base? In this paper we introduce an alternative to the usual model for user personalization.
Preventing Misleading Presentations of XML Documents: Some Initial Proposals BIBAFull-Text 492-496
  Alison Cawsey; Euan Dempster; Diana Bental; Daniel Pacey; Howard Williams
It is now straightforward to develop a range of different stylesheets to present XML documents in different ways, for example to create personalised presentations. If the XML document is available on the World Wide Web, then (subject to copyright) anyone can create their own stylesheet to present that document in new ways. This has the potential to allow improved "added value" services, such as personalised news feeders. But the power of stylesheet languages such as XSLT means that the document may be substantially transformed, with sections deleted or re-ordered. This re-structuring may result in a misleading and even dangerous presentation. This paper presents some proposals for putting some control in the hands of document authors, to allow them to indicate allowable transformations, and to provide a limited validation mechanism to verify that a transformed document meets the requirements of the author.
Adaptivity Conditions Evaluation for the User of Hypermedia Presentations Built with AHA! BIBAFull-Text 497-500
  Alessandra Cini; José Valdeni de Lima
The importance of adaptive hypermedia propelled the development of several authoring tools. Even though most tools are efficient for their designed task, they do not guarantee that the adaptivity is used in a high degree, in order to satisfy users with distinct goals, preferences, background and/or needs, as usual on the Web. In this paper we will present a system in development built to evaluate the adaptivity degree of presentations built into the AHA! authoring tool. The author, after build his presentation in AHA!, submits it to our system for evaluation. The results obtained from the evaluation can aid the author to know if the built presentation it is according to the proposed goals or the adaptivity degree should be increased.
Development of Adaptive Web Sites with Usability and Accessibility Features BIBAFull-Text 501-504
  Marta Fernández de Arriba; José A. López Brugos
This paper discusses a development system of Web sites adapted to the preferences of each user. A markup language (AWUML -- Adaptive Web site Usable Markup Language) based on XML technology is defined for use in the specification of the page contents, which guarantees the fulfillment of usability and accessibility requirements. Each user visualizes his own version of the web site resulting from the application of his profile, as much in presentation aspects as in content topics. The separation of document contents from its presentation makes maintenance of the web site easier. The use of this system facilitates access to the network, independent of the disabilities or the technical equipment limitations of the end user.
An Adaptive e-Commerce System Definition BIBAKFull-Text 505-509
  Francisco José García; Fabio Paternò; Ana Belén Gil
The emergence and growing popularity of electronic commerce and more specifically of catalogue-based sales, has made it a necessity to define flexible and adaptive e-commerce systems and architectures. Our work describes an e-commerce system that is suitable for small and medium enterprises, where several enterprises share a common e-commerce site, which automatically adapts its interface to offer end-users the products organised in electronic catalogues (e-catalogues). To be precise, in this paper we present the overall architecture of the e-commerce system, an architecture that is based on an agent-oriented technology. We also describe the actual state our system, which is called e-CoUSAL, and implements an adaptive agent in the server side of the architecture.
Keywords: Adaptive agent; E-commerce; Adaptivity in E-commerce; Adaptive interface; E-catalogue; XML
Web Site Personalization Using User Profile Information BIBAFull-Text 510-513
  Mohit Goel; Sudeshna Sarkar
In this paper we discuss a technique for web site personalization. Connectivity analysis has been shown to be useful in identifying high quality web pages within a topic or domain specific graph of hyper linked documents. We have implemented a system that creates a view of a subset of a web site most relevant to a given user. This sort of personalization is useful for filtering a useful subset of a site so that the user gets a low volume of quality information. The essence of our approach is to augment a previous connectivity analysis with content analysis. We present an agent which assists the user when he browses and distills a personalized sub graph of the website based on his user profile.
Who Do You Want to Be Today? Web Personae for Personalised Information Access BIBAFull-Text 514-517
  J. P. McGowan; Nicholas Kushmerick; Barry Smyth
Personalised context sensitivity is the Holy Grail of web information retrieval. As a first step towards this goal, we present the Web Personae personalised search and browsing system. We use well-known information retrieval techniques to develop and track user models. Web Personae differ from previous approaches in that we model users with multiple profiles, each corresponding to a distinct topic or domain. Such functionality is essential in heterogeneous environments such as the Web. We introduce Web Personae, describe an algorithm for learning such models from browsing data, and discuss applications and evaluation methods.
Adaptive Navigation Path Previewing for Learning on the Web BIBAFull-Text 518-521
  Akihiro Kashihara; Shinobu Hasegawa; Jun'ichi Toyoda
The main issue addressed in this paper is how to help learners plan a navigation path in existing web-based learning resources, which is an important process of self-directed learning in hyperspace. Our approach to this issue is to provide learners with the adaptive preview of a sequence of web pages as navigation path. Following the idea of path previewing, we have developed an assistant system. The system displays an overview of a web page selected by learners from a hyperspace map, by extracting information from the HTML document file related to the navigation path-planning context. It also enables learners to transform a sequence of previewed pages into a navigation path plan.
A Case-Based Recommender System Using Implicit Rating Techniques BIBAFull-Text 522-526
  Youngji Kim; Sooho Ok; Yongtae Woo
We propose a new case-based recommender system using implicit rating information. We present intra-attribute and inter-attribute weight derived from past interests of a user stored in the access logs, and a new similarity function to estimate similarities between new items set and the user profile. To verify the efficiency of our system, we have performed experimental comparisons between the proposed model and the collaborative filtering technique by mean absolute error (MAE) and receiver operating characteristic (ROC). The results show that the proposed model is more efficient than the traditional collaborative filtering technique.
IMAP -- Intelligent Multimedia Authoring Tools for Electronic Publishing BIBAFull-Text 527-530
  Sarit Kraus; Alexander Kröner; Lea Tsaban
IMAP provides software tools that support the authoring of electronic presentation by helping the author in multimedia content selection and layout design. IMAP consists of a Content Manager and a Layout Manager. In this paper we present the usage of IMAP tools for the authoring of on-line newspapers. The content selection is based both on the author's specifications and the user's interests, where the interrelations between objects play an important role in the evaluation of the set of objects. Layout management relies on a set of layout requirements taken from a layout profile and an author-defined style sheet. As was demonstrated in our experiments, integrating of both techniques yields interesting newspapers, whose layout can be customized by the reader.
A Hybrid Recommender System Combining Collaborative Filtering with Neural Network BIBAFull-Text 531-534
  Meehee Lee; Pyungseok Choi; Yongtae Woo
We propose a new recommender system which combines collaborative filtering(CF) with Self-Organizing Map (SOM) neural network. First, all users are segmented by demographic characteristics and users in each segment are clustered according to the preference of items using the SOM neural network. To recommend items to a user, CF algorithm is then applied on the cluster where the user belongs. As a result of experimentation for well-known movies, we show that the proposed system satisfies the predictability of CF algorithm in GroupLens. Also, our system improves the scalability and the performance of the traditional CF technique.
An Adaptive Cooperative Web Authoring Environment BIBAKFull-Text 535-538
  Ana María Martínez-Enríquez; Dominique Decouchant; Alberto L. Morán; Jesus Favela
Using AllianceWeb, authors distributed around the world can cooperate producing large documents in a consistent and concerted way. In this paper, we highlight the main aspects of the group awareness function that allows each author to diffuse his contribution to other co-authors, and to control the way by which other contributions are integrated into his environment. In order to support this function, essential to every groupware application, we have designed a self-adaptive cooperative interaction environment, parametrized by user preferences. Thus, the characteristics of an adaptive group awareness agent are defined.
Keywords: Cooperative and distributed authoring; World Wide Web; group awareness agent; GAA; adaptive cooperation interface
Adapting Learner Evaluation by Plans BIBAFull-Text 539-542
  Alke Martens; Adelinde M. Uhrmacher
State of the art intelligent tutoring systems (ITS) provide ample opportunity for adaptability. Typically, they adapt single pages and contents to the learner's knowledge level and evaluate the learner based on his past and current behavior. However, to decide whether a learner behaves in a coherent manner, the learning system requires some means for anticipating the learner's goals and intentions. Thus, not only the structure of the tutoring process, but also the evaluation strategy should be adaptive. The ITS PLAIT reasons about and evaluates the learner's behavior according to a plan adapted for that purpose. To facilitate the adaptation, the plan is structured into a three-layered hierarchy. The topmost layer reflects the teaching objectives suggested by pedagogy and the psychology of learning; the middle one embeds domain-dependent refinement. The third layer represents the steps the learner has taken and should optimally take.
WETAS: A Web-Based Authoring System for Constraint-Based ITS BIBAFull-Text 543-546
  Brent Martin; Antonija Mitrovic
Constraint-Based Modelling (CBM) is a student modelling technique for Intelligent Tutoring Systems (ITS) that is rapidly maturing. We have implemented several tutors using CBM, and demonstrated their suitability to, in particular, open-ended domains. It is easier to build tutors in some domains (e.g. open-ended) using CBM than other common approaches. We present WETAS (Web-Enabled Tutor Authoring System), a tutoring engine that facilitates the rapid implementation of ITS in new domains. We describe the architecture of WETAS, and give examples of two domains we have implemented.
Open-Ended Adaptive System for Facilitating Knowledge Construction in Web-Based Exploratory Learning BIBAFull-Text 547-550
  Hiroyuki Mitsuhara; Youji Ochi; Yoneo Yano
A typical learning method using the Internet is exploratory learning, where learners construct knowledge through exploring the web autonomously. A problem with web-based exploratory learning is insufficient hyperlinks. An open-ended adaptive system has been developed that facilitates knowledge construction in web-based exploratory learning by avoiding the impasse.
Adapting Web Interfaces by WHAT BIBAFull-Text 551-555
  Wilfred Ng; Jerry Yau
We have previously presented a Java database application, called WHAT, which is developed to analyse the users' navigation behaviour. Herein, we discuss our on-going work, which makes use the output results from WHAT to develop an enhanced interface for a commercial web site. The adapted interface provides the additional frames of popular trails and pages, which help reduce the navigation problem and facilitate better web topology adaptation.
Considering Sensing-Intuitive Dimension to Exposition-Exemplification in Adaptive Sequencing BIBAFull-Text 556-559
  Pedro Paredes; Pilar Rodriguez
This paper shows a way of using sensing-intuitive dimension of learning styles of students in order to improve the efficiency of adaptive learning systems. Firstly, it introduces the procedure of extracting information about sensing-intuitive students from the Felder-Soloman ILS questionnaire. Then, it presents a mechanism of application of sensing-intuitive dimension to exposition-exemplification sequencing. The example used to explain the adaptation effects is taken from a chess course developed with TANGOW, Task-based Adaptive learNer Guidance On the Web.
PlanEdit: An Adaptive Problem Solving Tool for Design BIBAFull-Text 560-563
  M. A. Redondo; C. Bravo; M. Ortega; M. F. Verdejo
Experimental learning environments based on simulation usually require monitoring and adaptation to the actions the user carries out. In this work we present the monitoring and adaptive features of PlanEdit, a tool for learning integral automation methods in buildings and housing by design and simulation.
Dynamic Generation of Interactive Dialogs Based on Intelligent Agents BIBAFull-Text 564-567
  Martín González Rodríguez; Benjamin López Pérez; María del Puerto Paule Ruíz
The design of the lexical and syntactical levels of the user interface of any web based system is commonly based on general human-computer interaction principles targeted to match the interaction requirements of the so-called typical user, an abstract generalization of each user of an application. However the identification of such typical user at the web design stage seems to be an impossible task, considering the dynamic nature of the web, where the user interface of a popular web systems can be used by thousands of different users everyday. In order to avoid this problem, we have designed GADEA, an intelligent user interface management system able to detect different kinds of users by mean of distributed data gathering agents. The information obtained by those agents is used to design interactive dialogues at execution time, which are adapted to the specific cognitive, perceptive and motor characteristics of the current user.
A Simple Architecture for Adaptive Hypertext BIBAFull-Text 568-571
  Oreste Signore
We describe a general, flexible architecture based on web standards (W3C Recommendations) where users access data through the mediation of intelligent agents. Architecture is based on simple but effective document and user models, with weighted and semantic tagging of both documents and links. User's profile and preferences affect the agents' behaviour, so that actions can occur on the server as well at the browser side. Profile and preferences accompany the URL reference, avoiding dependence from any central site.
Adaptive Learning Environment for Teaching and Learning in WINDS BIBAFull-Text 572-575
  Marcus Specht; Milos Kravcik; Roland Klemke; Leonid Pesin; Rüdiger Hüttenhain
This paper introduces one approach to e-Learning describing the Adaptive Learning Environment (ALE). The system provides a new methodological approach to design education on the web. ALE will be used to build a large knowledge base supporting Architecture and Civil Engineering Design Courses and to experiment a comprehensive Virtual University of Architecture and Engineering Design in the project WINDS. Here we outline the system architecture and present its learning environment. The system combines classical structuring of learning materials based on reusable learning objects with an alternative structure -- the course index.
Adaptation through Interaction Modelling in Educational Hypermedia BIBAFull-Text 576-579
  Elio Toppano
The poster describes MMforTED, an adaptive educational hypermedia system supporting the acquisition of several conceptualisations that can be used for reasoning about technical artefacts from different perspectives. Adaptation is performed by a next button that leads the learner to the best page to be read next (direct guidance). The decision which page is best is taken by analysing the properties of the course of interaction of the student up to a given time and selecting the next page on the base of the affordances of potential interaction it offers.
The Hybrid Model for Adaptive Educational Hypermedia BIBAFull-Text 580-585
  Mohamed Ramzy Zakaria; Adam Moore; Helen Ashman; Craig Stewart; Tim Brailsford
Web-based distance learning is becoming increasingly prevalent as the Internet permeates every aspect of our culture, and many educational content management systems are now in use on the web. However, learners' experiences of these systems are almost invariably static, with information being delivered regardless of their background or knowledge. Due to variation between learners', it is suggested that these web-based distance-learning systems would benefit from the capability of adapting their content to meet individual needs. To effectively implement this adaptation of educational material, we require a user model that supplies the system with information about the learners using the system, such as their backgrounds, knowledge, interests and learning styles. This paper focuses on presenting a user model that combines the advantages of two techniques (overlay and stereotyping) in a way that provides the system with the ability to deliver information that is fully informed by the requirements of individual users.

Doctoral Consortium

Navigation Modelling in Adaptive Hypermedia BIBAFull-Text 586-591
  Peter Dolog; Mária Bieliková
In this paper we reflect the need for modelling in a systematic production of adaptive hypermedia applications. Proposed approach is based on the Unified Modelling Language (UML). State diagrams are used to model possible paths through hypertext. The user model expressed by a class diagram determines structural and behavioural features, which are used for specification of adaptations in states and transitions contained in state diagrams.
An Adaptive Agent Model for e-Commerce Architecture BIBAKFull-Text 592-597
  Ana B. Gil; Francisco García; Zahia Guessoum
For building easily e-commerce applications, adaptive agent models are useful. These systems therefore should be equipped with the ability to improve their performances themselves, working over the basis of building a model with the goals, preferences and knowledge of each individual user in the environment. This paper presents a new architecture to support sales in Web stores centered on business based on product catalogs. The main goal of this architecture is focused on endowing capability of adaptivity in several ways: contents, navigation and presentation, while analyzing the different ways to develop the adaptivity in e-commerce.
Keywords: Adaptive Agents; E-commerce; Negotiation; User Model
Personalized Adaptation to Device Characteristics BIBAFull-Text 598-602
  Eelco Herder; Betsy van Dijk
Device characteristics, such as screen size and means of interaction, and the context in which a device is used, seriously affect the user's mental representation of an information environment and its intended use. We hypothesize that user characteristics are valuable resources for determining which information is of interest in specific situations. Our project goal is to design mechanisms for adapting navigation support to device characteristics and its context of use, thereby considering that user goals and the resulting expected navigation behavior might be subject to change.
Users Modeling for Adaptive Call Centers BIBAFull-Text 603-607
  Ilaria Torre
The project described in this paper applies the principles of adaptivity to a "traditional" Call Center in order to support the Operator in the interaction with the Customer. The system uses the models of both the Customer and the Operator and builds up the stepwise answer through an adaptive workflow.
The Exploratory of Personal Assistants BIBAFull-Text 608-612
  Lim Tek Yong; Tang Enya Kong
This paper presents a perspective of using personal assistants in personalized systems, particularly in both understanding the user model and generating appropriate adaptation and response. The paper discusses the use of a personal assistant as a data input source and the expected stereotypes based on mind-sets. This is the framework of exploring a user-centered design approach to building up the intelligent personalized system.