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

Proceedings of User Modeling 2005 2003-07-24

Fullname:Proceedings of the 10th International Conference on User Modeling
Editors:Liliana Ardissono; Paul Brna; Antonija Mitrovic
Location:Edinburgh, Scotland
Dates:2005-Jul-24 to 2005-Jul-29
Publisher:Springer-Verlag
Series:Lecture Notes in Computer Science, 2005, Volume 3538
Standard No:ISBN: 978-3-540-27885-6 (Print) 978-3-540-31878-1 (Online); hcibib: UMAP05
Papers:78
Pages:529
Links:Online Proceedings | Conference Home Page
Summary:The papers presented within this volume represent current work in the exciting area of user modeling -- an area that promises much to a range of economic and socially beneficial activities. The potential is enormous, and the applications of the technologies that have been developed are increasingly ambitious and relevant to the needs of the 21st century. The editors hope you enjoy, and benefit from, reading the papers within these proceedings.
  1. Invited Talks
  2. Adaptive Hypermedia
  3. Affective Computing
  4. Data Mining for Personalization and Cross-Recommendation
  5. ITS and Adaptive Advice
  6. Modeling and Recognizing Human Activity
  7. Multimodality and Ubiquitous Computing
  8. Recommender Systems
  9. Student Modeling
  10. User Modeling and Interactive Systems
  11. Web Site Navigation Support
  12. Doctoral Consortium Papers

Invited Talks

User Modeling Meets Usability Goals BIBAFull-Text 1-3
  Anthony Jameson
It has long been recognized that systems based on user modeling and adaptivity are associated with a number of typical usability problems -- which sometimes outweigh the benefits of adaptation. This talk will show that the anticipation and prevention of usability side effects should form an essential part of the iterative design of user-adaptive systems, just as the consideration of medical side effects plays a key role in the development of new medications. This strategy requires a comprehensive understanding of the reasons for typical usability problems and of strategies for preventing them.
Hey, That's Personal! BIBAFull-Text 4
  Lorrie Faith Cranor
Personalized online commerce systems, context-aware pervasive computing applications, and other personalized computing systems often sacrifice privacy for added convenience or improved service. Some systems provide users with substantial benefits from personalization; other systems profile users to the primary benefit of the service provider. In many cases users are not fully informed about how their profile information will be used and are not given adequate opportunities to control the use of their personal information. If developers of personalized systems do not consider privacy issues in the design of their systems they risk building systems that are unable to comply with legal requirements in some jurisdictions. In addition, concerns about privacy may slow adoption of some personalized systems or prevent them from ever gaining acceptance. In this talk I will discuss the privacy risks associated with personalization systems and discuss a number of approaches to reducing those risks, including approaches to minimizing the amount of profile information associated with identified individuals and approaches to better informing users and giving them meaningful opportunities to control the user of their personal information.
Inhabited Models: Supporting Coherent Behavior in Online Systems BIBAFull-Text 5-8
  Thomas Erickson
A principal focus of user modeling has been on modeling individuals, the aim being to support the design of interactive systems that can fluidly adapt to their users' needs. In this talk I shift the focus from interactions between a human and a computer, to interactions amongst people that are mediated by a digital system. My interest has to do with how to design online systems that can support the blend of flexibility and coherence that characterizes face to face interaction. I describe my approach, which involves creating shared visualizations of people and their activities in online situations such as chats, presentations, and auctions. This kind of visualization -- which serves as a sort of inhabited model of an activity -- plays a number of roles in supporting group interaction that is both flexible and coherent.

Adaptive Hypermedia

Integrating Open User Modeling and Learning Content Management for the Semantic Web BIBAFull-Text 9-18
  Ronald Denaux; Vania Dimitrova; Lora Aroyo
The paper describes an ontology-based approach for integrating interactive user modeling and learning content management to deal with typical adaptation problems, such as cold start and dynamics of the user's knowledge, in the context of the Semantic Web. An integrated OntoAIMS system is presented and its viability discussed based on user studies. The work demonstrates some novel aspects, such as (a) ontological approach for integration of methods for eliciting and utilizing of user models; (b) improved adaptation functionality resulted from that integration, validated with real users; (c) support of interoperability and reusability of adaptive components.
Modeling Suppositions in Users' Arguments BIBAFull-Text 19-29
  Sarah George; Ingrid Zukerman; Michael Niemann
During conversation, people often make assumptions or suppositions that are not explicitly stated. Failure to identify these suppositions may lead to mis-communication. In this paper, we describe a procedure that postulates such suppositions in the context of the discourse interpretation mechanism of BIAS -- a Bayesian Interactive Argumentation System. When a belief mentioned in a user's discourse differs from that obtained in BIAS' user model, our procedure searches for suppositions that explain this belief, preferring suppositions that depart minimally from the beliefs in the user model. Once a set of suppositions has been selected, it can be presented to the user for validation. Our procedure was evaluated by means of a web-based trial. Our results show that the assumptions posited by BIAS are considered sensible by our trial subjects.
Generative Programming Driven by User Models BIBAFull-Text 30-39
  Mauro Marinilli; Alessandro Micarelli
This paper discusses the automatic generation of programs by adapting the construction process to the user currently interacting with the program. A class of such systems is investigated where such generation process is continuously repeated making the program design and implementation evolve according to user behaviour. By leveraging on existing technologies (software generation facilities, modelling languages, specific and general standard metamodels) an experimental proof of concept system that is able to generate itself while interacting with the user is introduced and tested. The findings are discussed and a general organization for this class of adaptive systems is briefly proposed and compared with existing literature.

Affective Computing

Data-Driven Refinement of a Probabilistic Model of User Affect BIBAFull-Text 40-49
  Cristina Conati; Heather Maclaren
We present further developments in our work on using data from real users to build a probabilistic model of user affect based on Dynamic Bayesian Networks (DBNs) and designed to detect multiple emotions. We present analysis and solutions for inaccuracies identified by a previous evaluation; refining the model's appraisals of events to reflect more closely those of real users. Our findings lead us to challenge previously made assumptions and produce insights into directions for further improvement.
Recognizing Emotion from Postures: Cross-Cultural Differences in User Modeling BIBAKFull-Text 50-59
  Andrea Kleinsmith; P. Ravindra De Silva; Nadia Bianchi-Berthouze
The conveyance and recognition of human emotion and affective expression is influenced by many factors, including culture. Within the area of user modeling, it has become increasingly necessary to understand the role affect can play in personalizing interactive interfaces using embodied animated agents. Currently, little research focuses on the importance of emotion expression through body posture. Furthermore, little research aims at understanding cultural differences within this vein. Therefore, our goal is to evaluate whether or not differences exist in the way various cultures perceive emotion from body posture. We used images of 3D affectively expressive avatars to conduct recognition experiments with subjects from 3 cultures. The subjects' judgments were analyzed using multivariate analysis. We grounded the identified differences into a set of low-level posture features. Our results could prove useful for constructing affective posture recognition systems in cross-cultural environments.
Keywords: Affective communication; affective body postures; embodied animated agents; intercultural differences; user modeling
Recognizing, Modeling, and Responding to Users' Affective States BIBAFull-Text 60-69
  Helmut Prendinger; Junichiro Mori; Mitsuru Ishizuka
We describe a system that recognizes physiological data of users in real-time, interprets this information as affective states, and responds to affect by employing an animated agent. The agent assumes the role of an Empathic Companion in a virtual job interview scenario where it accompanies a human interviewee. While previously obtained results with the companion with were not significant, the analysis reported here demonstrates that empathic feedback of an agent may reduce user arousal while hearing interviewer questions. This outcome may prove useful for educational systems or applications that induce user stress.
Using Learner Focus of Attention to Detect Learner Motivation Factors BIBAFull-Text 70-73
  Lei Qu; Ning Wang; W. Lewis Johnson
This paper presents a model for pedagogical agents to use the learner's attention to detect motivation factors of the learner in interactive learning environments. This model is based on observations from human tutors coaching students in on-line learning tasks. It takes into account the learner's focus of attention, current task, and expected time required to perform the task. A Bayesian model is used to combine evidence from the learner's eye gaze and interface actions to infer the learner's focus of attention. Then the focus of attention is combined with information about the learner's activities, inferred by a plan recognizer, to detect the learner's degree of confidence, confusion and effort. Finally, we discuss the results of an empirical study that we performed to evaluate our model.
Player Modeling Impact on Player's Entertainment in Computer Games BIBAFull-Text 74-78
  Georgios N. Yannakakis; Manolis Maragoudakis
In this paper we introduce an effective mechanism for obtaining computer games of high interest (i.e. satisfaction for the player). The proposed approach is based on the interaction of a player modeling tool and a successful on-line learning mechanism from the authors' previous work on prey/predator computer games. The methodology demonstrates high adaptability into dynamical playing strategies as well as reliability and justifiability to the game user.

Data Mining for Personalization and Cross-Recommendation

Using Learning Curves to Mine Student Models BIBAFull-Text 79-88
  Brent Martin; Antonija Mitrovic
This paper presents an evaluation study that measures the effect of modifying feedback generality in an Intelligent Tutoring System (ITS) based on Student Models. A taxonomy of the tutor domain was used to group existing knowledge elements into plausible, more general, concepts. Existing student models were then used to measure the validity of these new concepts, demonstrating that at least some of these concepts appear to be more effective at capturing what the students learned than the original knowledge elements. We then trialled an experimental ITS that gave feedback at a higher level. The results suggest that it is feasible to use this approach to determine how feedback might be fine-tuned to better suit student learning, and hence that learning curves are a useful tool for mining student models.
Exploiting Probabilistic Latent Information for the Construction of Community Web Directories BIBAFull-Text 89-98
  Dimitrios Pierrakos; Georgios Paliouras
This paper improves a recently-presented approach to Web Personalization, named Community Web Directories, which applies personalization techniques to Web Directories. The Web directory is viewed as a concept hierarchy and personalization is realized by constructing user community models on the basis of usage data collected by the proxy servers of an Internet Service Provider. The user communities are modeled using Probabilistic Latent Semantic Analysis (PLSA), which provides a number of advantages such as overlapping communities, as well as a good rationale for the associations that exist in the data. The data that are analyzed present challenging peculiarities such as their large volume and semantic diversity. Initial results presented in this paper illustrate the effectiveness of the new method.
ExpertiseNet: Relational and Evolutionary Expert Modeling BIBAFull-Text 99-108
  Xiaodan Song; Belle L. Tseng; Ching-Yung Lin; Ming-Ting Sun
We develop a novel user-centric modeling technology, which can dynamically describe and update a person's expertise profile. In an enterprise environment, the technology can enhance employees' collaboration and productivity by assisting in finding experts, training employees, etc. Instead of using the traditional search methods, such as the keyword match, we propose to use relational and evolutionary graph models, which we call ExpertiseNet, to describe and find experts. These ExpertiseNets are used for mining, retrieval, and visualization. We conduct experiments by building ExpertiseNets for researchers from a research paper collection. The experiments demonstrate that expertise mining and matching are more efficiently achieved based on the proposed relational and evolutionary graph models.
Task-Oriented Web User Modeling for Recommendation BIBAFull-Text 109-118
  Xin Jin; Yanzan Zhou; Bamshad Mobasher
We propose an approach for modeling the navigational behavior of Web users based on task-level patterns. The discovered "tasks" are characterized probabilistically as latent variables, and represent the underlying interests or intended navigational goal of users. The ability to measure the probabilities by which pages in user sessions are associated with various tasks, allow us to track task transitions and modality shifts within (or across) user sessions, and to generate task-level navigational patterns. We also propose a maximum entropy recommendation system which combines the page-level statistics about users' navigational activities together with our task-level usage patterns. Our experiments show that the task-level patterns provide better interpretability of Web users' navigation, and improve the accuracy of recommendations.
Ontologically-Enriched Unified User Modeling for Cross-System Personalization BIBAFull-Text 119-123
  Bhaskar Mehta; Claudia Niederee; Avare Stewart; Marco Degemmis; Pasquale Lops
Personalization today has wide spread use on many Web sites. Systems and applications store preferences and information about users in order to provide personalized access. However, these systems store user profiles in proprietary formats. Although some of these systems store similar information about the user, exchange or reuse of information is not possible and information is duplicated. Additionally, since user profiles tend to be deeply buried inside such systems, users have little control over them. This paper proposes the use of a common ontology-based user context model as a basis for the exchange of user profiles between multiple systems and, thus, as a foundation for cross-system personalization.

ITS and Adaptive Advice

Using Student and Group Models to Support Teachers in Web-Based Distance Education BIBAFull-Text 124-133
  Essam Kosba; Vania Dimitrova; Roger Boyle
The paper illustrates how student modeling and advice generation methods can be used to address problems experienced in Web-based distance education courses. We have developed the TADV (Teacher ADVisor) framework which builds student models based on the tracking data collected by a course management system and uses these models to generate advice to the course instructors, so that they can improve their feedback and guidance to distance students. The paper introduces TADV, describes how student, group, and class models are used for generating advice to the teachers, and discusses the viability of this approach based on an evaluative study with users.
Using Similarity to Infer Meta-cognitive Behaviors During Analogical Problem Solving BIBAFull-Text 134-143
  Kasia Muldner; Cristina Conati
We present a computational framework designed to provide adaptive support aimed at triggering learning from problem-solving activities in the presence of worked-out examples. The key to the framework's ability to provide this support is a user model that exploits a novel classification of similarity to infer the impact of a particular example on a given student's metacognitive behaviors and subsequent learning.
COPPER: Modeling User Linguistic Production Competence in an Adaptive Collaborative Environment BIBAFull-Text 144-153
  Timothy Read; Elena Bárcena; Beatriz Barros; Raquel Varela; Jesús Pancorbo
This article starts from the standard conceptualization of linguistic competence as being composed of four related memories of comparable relevance: reading, listening, writing and speaking. It is argued that there is a considerable imbalance between the application of technology to the former two and the others. A system called COPPER is presented, which addresses this problem by helping students to improve their linguistic production combining individual and collaborative activities in a constructivist methodology with a way to overcome technological language analysis difficulties. The knowledge models used in COPPER have been developed from the authors' previous work, undertaken to solve some of the problems of linguistic models of student competence. Methodologically, the system 'empowers' students in that it leads to shared understanding, which reinforces learning. The system is adaptive in the sense that group formation is dynamic and based upon the nature of the tasks to be performed and the features of the student model.
User Cognitive Style and Interface Design for Personal, Adaptive Learning. What to Model? BIBAFull-Text 154-163
  Elizabeth Uruchrutu; Lachlan MacKinnon; Roger Rist
The concept of personal learning environments has become a significant research topic over the past few years. Building such personal, adaptive environments requires the convergence of several modeling dimensions and an interaction strategy based on a user model that incorporates key cognitive characteristics of the learners. This paper reports on an initial study carried out to evaluate the extent to which matching the interface design to the learner cognitive style facilitates learning performance. Results show that individual differences influence the way learners react to and perform under different interface conditions, however no simple effects were observed that confirm a relationship between cognitive style and interface affect.
Tailored Responses for Decision Support BIBAFull-Text 164-168
  Terrence Harvey; Sandra Carberry; Keith Decker
Individuals differ in the resources that they are willing to expend on information gathering and on the importance of different kinds of information. We have developed MADSUM, a system that takes into account user constraints on resources, the significance of propositions that might be included in a response, and the user's priorities with respect to resource and content attributes; MADSUM produces a response tailored to individual users in a decision support setting.
Decision Theoretic Dialogue Planning for Initiative Problems BIBAFull-Text 169-173
  Bryan McEleney; Gregory O'Hare
The taking of initiative has significance in spoken language dialogue systems and in human-computer interaction. A system that takes no initiative may fail to seize opportunities that are important, but a system that always takes the initiative may not allow the user to take the actions he favours. We have implemented a mixed-initiative planning system that adapts its strategy to a nested belief model. In simulation, the planner's performance was compared to two fixed strategies of always taking the initiative and always declining it, and it performed significantly better than both.
A Semi-automated Wizard of Oz Interface for Modeling Tutorial Strategies BIBAFull-Text 174-178
  Paola Rizzo; Hyokyeong Lee; Erin Shaw; W. Lewis Johnson; Ning Wang; Richard E. Mayer
Human teaching strategies are usually inferred from transcripts of face-to-face conversations or computer-mediated dialogs between learner and tutor. However, during natural interactions there are no constraints on the human tutor's behavior and thus tutorial strategies are difficult to analyze and reproduce in a computational model. To overcome this problem, we have realized a Wizard of Oz interface, which by constraining the tutor's interaction makes explicit his decisions about why, how, and when to assist the student in a computer-based learning environment. These decisions automatically generate natural language utterances of different types according to two "politeness" strategies. We have successfully used the interface to model tutorial strategies.

Modeling and Recognizing Human Activity

Generating Artificial Corpora for Plan Recognition BIBAFull-Text 179-188
  Nate Blaylock; James Allen
Corpora for training plan recognizers are scarce and difficult to gather from humans. However, corpora could be a boon to plan recognition research, providing a platform to train and test individual recognizers, as well as allow different recognizers to be compared. We present a novel method for generating artificial corpora for plan recognition. The method uses a modified AI planner and Monte-Carlo sampling to generate action sequences labeled with their goal and plan. This general method can be ported to allow the automatic generation of corpora for different domains.
Reasoning About Interaction in a Multi-user System BIBAFull-Text 189-198
  Michael Y. K. Cheng; Robin Cohen
This paper presents a model for an agent to reason about interaction with multiple users in a collaborative environment. Central to this model is the concept of an interaction strategy, determining both who to ask and what to ask, towards maximizing overall expected utility. We allow for the case of a user not responding at all, after a period of waiting, and a user responding "I don't know". Our model determines how long to wait for a response, and provides for follow up questions to users. All of this is done in a user modeling approach, with decisions based on specific factors being modeled for each user. We present the model in detail, using examples to illustrate its effectiveness and contrasting with related work.
A Comparison of HMMs and Dynamic Bayesian Networks for Recognizing Office Activities BIBAFull-Text 199-209
  Nuria Oliver; Eric Horvitz
We present a comparative analysis of a layered architecture of Hidden Markov Models (HMMs) and dynamic Bayesian networks (DBNs) for identifying human activities from multimodal sensor information. We use the two representations to diagnose users' activities in S-SEER, a multimodal system for recognizing office activity from real-time streams of evidence from video, audio and computer (keyboard and mouse) interactions. As the computation required for sensing and processing perceptual information can impose significant burdens on personal computers, the system is designed to perform selective perception using expected-value-of-information (EVI) to limit sensing and analysis. We discuss the relative performance of HMMs and DBNs in the context of diagnosis and EVI computation.
Modeling Agents That Exhibit Variable Performance in a Collaborative Setting BIBAFull-Text 210-219
  Ingrid Zukerman; Christian Guttmann
In a collaborative environment, knowledge about collaborators' skills is an important factor when determining which team members should perform a task. However, this knowledge may be incomplete or uncertain. In this paper, we extend our ETAPP (Environment-Task-Agents-Policy-Protocol) collaboration framework by modeling team members that exhibit non-deterministic performance, and comparing two alternative ways of using these models to assign agents to tasks. Our simulation-based evaluation shows that performance variability has a large impact on task performance, and that task performance is improved by consulting agent models built from a small number of observations of agents' recent performance.
Detecting When Students Game the System, Across Tutor Subjects and Classroom Cohorts BIBAFull-Text 220-224
  Ryan Shaun Baker; Albert T. Corbett; Kenneth R. Koedinger; Ido Roll
Building a generalizable detector of student behavior within intelligent tutoring systems presents two challenges: transferring between different cohorts of students (who may develop idiosyncratic strategies of use), and transferring between different tutor lessons (which may have considerable variation in their interfaces, making cognitively equivalent behaviors appear quite different within log files). In this paper, we present a machine-learned detector which identifies students who are "gaming the system", attempting to complete problems with minimal cognitive effort, and determine that the detector transfers successfully across student cohorts but less successfully across tutor lessons.
A Bayesian Approach to Modelling Users' Information Display Preferences BIBAFull-Text 225-230
  Beate Grawemeyer; Richard Cox
This paper describes the process by which we constructed a user model for ERST -- an External Representation Selection Tutor -- which recommends external representations (ERs) for particular database query task types based upon individual preferences, in order to enhance ER reasoning performance. The user model is based on experimental studies which examined the effect of background knowledge of ERs upon performance and preferences over different types of tasks.
Modeling of the Residual Capability for People with Severe Motor Disabilities: Analysis of Hand Posture BIBAFull-Text 231-235
  Rachid Kadouche; Mounir Mokhtari; Marc Maier
People with severe motor disabilities use mainly their residual motor capability for the use of technical aids, and for the control of input devices to technical aids. This paper describes our work on characterizing the motor capability of the upper arm for patients with severe motor disabilities. This work is a continuation of a project aimed at modeling the arm posture of quadriplegic patients using STS (Spatial Tracking System) and at analyzing the compensatory strategies developed by hemiplegic patients while accessing physical interfaces for technical aids [5]. Here we report work undertaken for analyzing the posture of the hand: we have developed two calibration methods for the Cyberglove and compare their utility and ergonomics in applications on patients with motor disabilities. The first type of calibration proceeds sequentially and takes into account one joint after the other (of the hand and each digit), whereas the second procedure is based on a few key postures calibrating several joints at once. To compare the precision of both methods, four healthy subjects participated in experiments using the Cyberglove. We show that the first type of calibration is more accurate but takes longer, whereas the second is less accurate but shorter. This trade-off might be acceptable for assessing the manual workspace in patients with motor disabilities. In particular, excessive muscular fatigue and limited dexterity are decisive factors for choosing the calibration by key postures in patients. We applied the calibration by key postures to three myopathic patients and individually quantified their restricted manual working space.
Non-intrusive User Modeling for a Multimedia Museum Visitors Guide System BIBAFull-Text 236-240
  Tsvi Kuflik; Charles Callaway; Dina Goren-Bar; Cesare Rocchi; Oliviero Stock
A personalized multimedia museum visitor's guide system may be a valuable tool for improving user satisfaction in a museum visit. Personalization poses challenges to user modeling in the museum environment, especially when several different applications are supported by the same user model, where it is required to operate in a non-intrusive manner. This work presents the PEACH experience of non-intrusive user modeling supporting online dynamic multimedia presentation production and additional applications such as visit summary report generation.

Multimodality and Ubiquitous Computing

Modelling the Behaviour of Elderly People as a Means of Monitoring Well Being BIBAFull-Text 241-250
  Nick Hine; Andrew Judson; Saqib Ashraf; John Arnott; Andrew Sixsmith; Steve Brown
The care of elderly people in their own homes is being promoted throughout the world. The proportion of older people within western societies is rising, and it is anticipated that the already stretched resources of both the informal and formal care sectors will be unable to meet demand for home based care in the near future. This paper reports on work being undertaken within the BT Care in the Community project to model the lives of older people in order to understand, anticipate and respond to their home based care needs.
Bayesphone: Precomputation of Context-Sensitive Policies for Inquiry and Action in Mobile Devices BIBAFull-Text 251-260
  Eric Horvitz; Paul Koch; Raman Sarin; Johnson Apacible; Muru Subramani
Inference and decision making with probabilistic user models may be infeasible on portable devices such as cell phones. We highlight the opportunity for storing and using precomputed inferences about ideal actions for future situations, based on offline learning and reasoning with the user models. As a motivating example, we focus on the use precomputation of call-handling policies for cell phones. The methods hinge on the learning of Bayesian user models for predicting whether users will attend meetings on their calendar and the cost of being interrupted by incoming calls should a meeting be attended.
Just Do What I Tell You: The Limited Impact of Instructions on Multimodal Integration Patterns BIBAFull-Text 261-270
  Sharon Oviatt; Rachel Coulston; Rebecca Lunsford
Large individual differences have been documented among users in their multimodal integration patterns, which suggest that new user-adaptive approaches to multimodal fusion may be opportune. Before pursuing such an approach, this study explores whether people can be successfully encouraged to switch their multimodal integration pattern to one that is easier to process through the use of explicit instructions. Longitudinal data were collected from young and elderly adults as they used speech and pen input with a simulated map system. Results revealed that only 37% of users switched their integration pattern and maintained it, whereas another 19% never switched their natural pattern and 31% switched but then reverted during a follow-up session. In addition, significant destabilization of elderly users' integration pattern was one "cost" of attempting to instruct a change in pattern. This research underscores the need for user-centered design in future multimodal system development, especially for vulnerable users such as the elderly.
Motion-Based Adaptation of Information Services for Mobile Users BIBAFull-Text 271-276
  Mathias Bauer; Matthieu Deru
Adaptive information systems typically exploit knowledge about the user's interests, preferences, goals etc. to determine what should be presented to the user and how this presentation should take place. When dealing with mobile users, however, information about their motions -- the places visited, the duration of stays, average velocity etc. -- can be additionally exploited to enrich the user model and better adapt the system behavior to the user's needs. This paper discusses the use of positioning data and background knowledge to achieve such a motion-based adaptation of information provision.
Interaction-Based Adaptation for Small Screen Devices BIBAFull-Text 277-281
  Enrico Bertini; Andrea Calì; Tiziana Catarci; Silvia Gabrielli; Stephen Kimani
This paper explores an original approach to overcome current issues in the use of mobile devices, such as limited screen space and interaction modalities, based on exploiting interface adaptation and adaptive techniques. Specifically, the paper describes the application of this approach to a web searching prototype, which collects usage data to model interaction and provide a personalized version of the web facility visited by the user.
Adapting Home Behavior to Its Inhabitants BIBAFull-Text 282-286
  Berardina De Carolis
In this paper, we propose a multiagent system for simulating the control of an intelligent home able to adapt its behavior to the user situation. Central to the adaptation process is the concept of influence sphere that is defined in function of the type of service it provides to house inhabitants (i.e. comfort, security, entertainment, etc.). Each influence sphere is controlled by a Supervisor Agent (SA) that is responsible for taking decisions relative to that scope. Decisions about actions involve device behaviors that, in our system, are controlled by Operator Agents (OAs). Each OA is responsible for deciding the utility of an action in the current user context. Then, according to this organization, the adaptation process is performed at two levels: globally for the relevant influence sphere and locally at the device level.

Recommender Systems

Design and Evaluation of a Music Retrieval Scheme That Adapts to the User's Impressions BIBAFull-Text 287-296
  Tadahiko Kumamoto
We have developed a scheme for music retrieval that adapts to the user's impressions of the musical pieces. First, we conducted impression-estimation experiments in which 100 subjects gave their impression of 80 musical pieces, and then, using a clustering method, we classified the 100 subjects into 20 groups based on the results. Next, we created a user model for each group consisting of formulas for numerically expressing the impressions and a set of vectors calculated using the formulas. We then developed a procedure for identifying the most suitable model for an unidentified user. Testing of the models and procedure in an existing impression-based music-retrieval system demonstrated the effectiveness of the proposed scheme.
The Pursuit of Satisfaction: Affective State in Group Recommender Systems BIBAFull-Text 297-306
  Judith Masthoff
This paper describes three algorithms to model and predict the satisfaction experienced by individuals using a group recommender system which recommends sequences of items. Satisfaction is treated as an affective state. In particular, we model the wearing off of emotion over time and assimilation effects, where the affective state produced by previous items influences the impact on satisfaction of the next item. We compare the algorithms with each other, and investigate the effect of parameter values by comparing the algorithms' predictions with the results of an earlier empirical study. We show a way in which affective state can be used in recommender systems, which is useful for recommendations not only to groups but also to individuals.
An Economic Model of User Rating in an Online Recommender System BIBAFull-Text 307-316
  F. Maxwell Harper; Xin Li; Yan Chen; Joseph A. Konstan
Economic modeling provides a formal mechanism to understand user incentives and behavior in online systems. In this paper we describe the process of building a parameterized economic model of user-contributed ratings in an online movie recommender system. We constructed a theoretical model to formalize our initial understanding of the system, and collected survey and behavioral data to calibrate an empirical model. This model explains 34% of the variation in user rating behavior. We found that while economic modeling in this domain requires an initial understanding of user behavior and access to an uncommonly broad set of user survey and behavioral data, it returns significant formal understanding of the activity being modeled.
Incorporating Confidence in a Naive Bayesian Classifier BIBAKFull-Text 317-326
  V. Pronk; S. V. R. Gutta; W. F. J. Verhaegh
Naive Bayes is a relatively simple classification method to, e.g., rate TV programs as interesting or uninteresting to a user. In case the training set consists of instances, chosen randomly from the instance space, the posterior probability estimates are random variables. Their statistical properties can be used to calculate confidence intervals around them, enabling more refined classification strategies than the usual argmax-operator. This may alleviate the cold-start problem and provide additional feedback to the user.
   In this paper, we give an explicit expression to estimate the variances of the posterior probability estimates from the training data and investigate the strategy that refrains from classification in case the confidence interval around the largest posterior probability overlaps with any of the other intervals.
   We show that the classification error rate can be significantly reduced at the cost of a lower coverage, i.e., the fraction of classifiable instances, in a TV-program recommender.
Keywords: machine learning; naive Bayes; recommenders; reliability; confidence intervals
Modeling User's Opinion Relevance to Recommending Research Papers BIBAKFull-Text 327-331
  Sílvio César Cazella; Luis Otávio Campos Alvares
Finding the right material on the Web could be a worthwhile result. Users waste too much time to discover the useful information. Recommender system can provide some shortcuts to the user, but if the recommendation is based on people's opinion, one question remains -- how relevant is a user's opinion? This paper presents a model to define the user's relevance opinion in a recommender system. This metric aims to help the target user to decide in what recommendation he should focus his attention. Beyond the model, we present a real experiment using an e-government database.
Keywords: User modeling; Authority; Recommender System
User- and Community-Adaptive Rewards Mechanism for Sustainable Online Community BIBAFull-Text 332-336
  Ran Cheng; Julita Vassileva
Abundance of user contributions does not necessarily indicate sustainability of an online community. On the contrary, excessive contributions in the systems may result in "information overload" and user withdrawal. We propose an adaptive rewards mechanism aiming to restrict the quantity of the contributions, elicit contributions with higher quality and simultaneously inhibit inferior ones. The mechanism adapts to the users preferences with respect to types of contributions and to the current needs of the community depending on the time and the number of existing contributions.
Off-line Evaluation of Recommendation Functions BIBAFull-Text 337-341
  Tingshao Zhu; Russ Greiner; Gerald Häubl; Kevin Jewell; Bob Price
This paper proposes a novel method for assessing the performance of any Web recommendation function (ie user model), M, used in a Web recommender system, based on an off-line computation using labeled session data. Each labeled session consists of a sequence of Web pages followed by a page p(S) that contains information the user claims is relevant. We then apply M to produce a corresponding suggested page p(S). In general, we say that M is good if p(S) has content "similar" to the associated p(IC), based on the same session. This paper defines a number of functions for estimating this p(S) to p(IC) similarity that can be used to evaluate any new models off-line, and provides empirical data to demonstrate that evaluations based on these similarity functions match our intuitions.
Evaluating the Intrusion Cost of Recommending in Recommender Systems BIBAFull-Text 342-346
  Felix Hernandez-del-Olmo; Elena Gaudioso; Jesus G. Boticario
Recommender systems suggest items, guiding the user in a personalized way in a large space of possible options. To accomplish this task, they should try to bother users as less as possible, but each recommendation occupies expensive room in the always small user interface. Unfortunately, current evaluation of recommender systems do not have into account this cost. This work presents some new measures that have into account this intrusion cost while recommending. Some experiments are performed to compare our approach with traditional ones.

Student Modeling

Introducing Prerequisite Relations in a Multi-layered Bayesian Student Model BIBAFull-Text 347-356
  C. Carmona; E. Millán; J. L. Pérez-de-la-Cruz; M. Trella; R. Conejo
In this paper we present an extension of a previously developed generic student model based on Bayesian Networks. A new layer has been added to the model to include prerequisite relationships. The need of this new layer is motivated from different points of view: in practice, this kind of relationships are very common in any educational setting, but also their use allows for improving efficiency of both adaptation mechanisms and the inference process. The new prerequisite layer has been evaluated using two different experiments: the first experiment uses a small toy example to show how the BN can emulate human reasoning in this context, while the second experiment with simulated students suggests that prerequisite relationships can improve the efficiency of the diagnosis process by allowing increased accuracy or reductions in the test length.
Exploring Eye Tracking to Increase Bandwidth in User Modeling BIBAFull-Text 357-366
  Cristina Conati; Christina Merten; Kasia Muldner; David Ternes
The accuracy of a user model usually depends on the amount and quality of information available on the user's states of interest. An eye-tracker provides data detailing where a user is looking during interaction with the system. In this paper we present a study to explore how this information can improve the performance of a model designed to assess the user's tendency to engage in a meta-cognitive behavior known as self-explanation.
Modeling Students' Metacognitive Errors in Two Intelligent Tutoring Systems BIBAFull-Text 367-376
  Ido Roll; Ryan S. Baker; Vincent Aleven; Bruce M. McLaren; Kenneth R. Koedinger
Intelligent tutoring systems help students acquire cognitive skills by tracing students' knowledge and providing relevant feedback. However, feedback that focuses only on the cognitive level might not be optimal -- errors are often the result of inappropriate metacognitive decisions. We have developed two models which detect aspects of student faulty metacognitive behavior: A prescriptive rational model aimed at improving help-seeking behavior, and a descriptive machine-learned model aimed at eliminating attempts to "game" the tutor. In a comparison between the two models we found that while both successfully identify gaming behavior, one is better at characterizing the types of problems students game in, and the other captures a larger variety of faulty behaviors. An analysis of students' actions in two different tutors suggests that the help-seeking model is domain independent, and that students' behavior is fairly consistent across classrooms, age groups, domains, and task elements.
Modeling Individual and Collaborative Problem Solving in Medical Problem-Based Learning BIBAFull-Text 377-386
  Siriwan Suebnukarn; Peter Haddawy
Since problem solving in group problem-based learning is a collaborative process, modeling individuals and the group is necessary if we wish to develop an intelligent tutoring system that can do things like focus the group discussion, promote collaboration, or suggest peer helpers. We have used Bayesian networks to model individual student knowledge and activity, as well as that of the group. The validity of the approach has been tested with student models in the areas of head injury, stroke and heart attack. Receiver operating characteristic (ROC) curve analysis shows that, the models are highly accurate in predicting individual student actions. Comparison with human tutors shows that group activity determined by the model agrees with that suggested by the majority of the human tutors with a high degree of statistical agreement (McNemar test, p = 0.774, Kappa = 0.823).
User Modeling in a Distributed E-Learning Architecture BIBAFull-Text 387-391
  Peter Brusilovsky; Sergey Sosnovsky; Olena Shcherbinina
This paper is focused on user modeling and adaptation in distributed E-Learning systems. We describe here CUMULATE, a generic student modeling server developed for a distributed E-Learning architecture, KnowledgeTree. We also introduce a specific, topic-based knowledge modeling approach which has been implemented as an inference agent in CUMULATE and used in QuizGuide, an adaptive system that helps students select the most relevant self-assessment quizzes. We also discuss our attempts to evaluate this multi-level student modeling.
Computer Adaptive Testing: Comparison of a Probabilistic Network Approach with Item Response Theory BIBAKFull-Text 392-396
  Michel C. Desmarais; Xiaoming Pu
Bayesian and probabilistic networks are claimed to offer powerful approaches to inferring an individual's knowledge state from evidence of mastery of concepts or skills. A typical application where such tools can be useful is Computer Adaptive Testing (CAT). Bayesian networks have been proposed as an alternative to the traditional Item Response Theory (IRT), which has been the prevalent CAT approach for the last three decades. We compare the performance of one probabilistic network approach, named POKS, to the IRT two parameter logistic model. Experimental results over a 34 items UNIX test and a 160 items French language test show that both approaches can classify examinees as master or non master effectively and efficiently. Implications of these results for adaptive testing and student modeling are discussed.
Keywords: CAT; IRT; Probabilistic networks; Bayesian networks; adaptive testing; student models; knowledge assessment
A Framework for Browsing, Manipulating and Maintaining Interoperable Learner Profiles BIBAFull-Text 397-401
  Peter Dolog; Michael Schäfer
Learners are assessed by several systems during their life-long learning. Those systems can maintain fragments of information about a learner derived from his learning performance and/or assessment in that particular system. Customization services would perform better if they would be able to exchange as many relevant fragments of information about the learner as possible. This paper presents the conceptualization and implementation of a framework which provides a common base for the exchange of learner profiles between several sources. The exchange representation of learner profiles is based on standards. An API is designed and implemented to create/export and manipulate such learner profiles. The API is implemented for two cases, as a Java API and as web services with synchronized model exchange between multiple sources. Application cases of the API are discussed shortly as well.
Towards Efficient Item Calibration in Adaptive Testing BIBAFull-Text 402-406
  Eduardo Guzmán; Ricardo Conejo
Reliable student models are vital for the correct functioning of Intelligent Tutoring Systems. This means that diagnosis tools used to update the student models must be also reliable. Through adaptive testing, student knowledge can be inferred. The tests are based on a psychometric theory, the Item Response Theory. In this theory, each question has a function assigned that is essential for determining student knowledge. These functions must be previously inferred by means of calibration techniques that use non-adaptive student test sessions. The problem is that, in general, calibration algorithms require huge sets of sessions. In this paper, we present an efficient calibration technique that just requires a reduced set of prior sessions.
Synergy of Performance-Based Model and Cognitive Trait Model in DP-ITS BIBAFull-Text 407-411
  Zoran Jeremic; Taiyu Lin; A Kinshuk; Vladan Devedzic
Information about the student in student model is the basis for virtual learning environments to provide the necessary adaptation. Cognitive Trait Model (CTM) profiles the student based on cognitive traits, such as his/her working memory capacity and inductive reasoning ability. Performance-based adaptation can guide the student to the required concept, whereas cognitive support serves to prevent the student's cognitive overload while still representing sufficient challenges to the student. This paper describes the synergy of a performance-based student model and CTM in an intelligent tutoring system called DP-ITS.
Up and Down the Number-Line: Modelling Collaboration in Contrasting School and Home Environments BIBAFull-Text 412-416
  Hilary Tunley; Benedict du Boulay; Rosemary Luckin; Joe Holmberg; Joshua Underwood
This paper is concerned with user modelling issues such as adaptive educational environments, adaptive information retrieval, and support for collaboration. The HomeWork project is examining the use of learner modelling strategies within both school and home environments for young children aged 5 -- 7 years. The learning experience within the home context can vary considerably from school especially for very young learners, and this project focuses on the use of modelling which can take into account the informality and potentially contrasting learning styles experienced within the home and school.

User Modeling and Interactive Systems

Temporal Blurring: A Privacy Model for OMS Users BIBAFull-Text 417-422
  Rosa A. Alarcón; Luis A. Guerrero; José A. Pino
Stereotypes and clustering are some techniques for creating user models from user behavior. Yet, they possess important risks as users actions could be misinterpreted or users could be associated with undesirable profiles. It could be worst if users' actions, beliefs, and comments are long term stored such as in Organizational Memory Systems (OMS) where users' contributions are available to the whole organization. We propose a privacy model based on four privacy roles that allow users to control the disclosure of their personal data and, when recovered, blurs such data as time passes.
A Framework of Context-Sensitive Visualization for User-Centered Interactive Systems BIBAKFull-Text 423-427
  Eui-Chul Jung; Keiichi Sato
This research proposes an adaptive mechanism of information visualizing that responds to context changes in knowledge-intensive work. A framework of Context-Sensitive Visualization (CSV) was introduced as a conceptual foundation for developing a middleware with three features to maximize performance of interactive systems. These features provide a mechanism for selecting appropriate content, scope, resolution, format, and timing of information delivery for effective use in changing context. In order to embed context sensitivity into the information mapping and visualization, the concept of the Context-Sensitive Object (CSO) was developed as a basic system structure for implementing the CSV.
Keywords: Context-Sensitive Visualization; Knowledge-Base; Interactive System; Context-Sensitive Object
Gumo -- The General User Model Ontology BIBAKFull-Text 428-432
  Dominik Heckmann; Tim Schwartz; Boris Brandherm; Michael Schmitz
We introduce the general user model ontology Gumo for the uniform interpretation of distributed user models in intelligent semantic web enriched environments. We discuss design decisions, show the relation to the user model markup language UserML and present the integration of ubiquitous applications with the u2m.org user model service.
Keywords: User model ontology; semantic web; ubiquitous user model service; intelligent environments; user model markup language
Balancing Awareness and Interruption: Investigation of Notification Deferral Policies BIBAFull-Text 433-437
  Eric Horvitz; Johnson Apacible; Muru Subramani
We review experiments with bounded deferral, a method aimed at reducing the disruptiveness of incoming messages and alerts in return for bounded delays in receiving information. Bounded deferral provides users with a means for balancing awareness about potentially urgent information with the cost of interruption.
A Decomposition Model for the Layered Evaluation of Interactive Adaptive Systems BIBAFull-Text 438-442
  Alexandros Paramythis; Stephan Weibelzahl
A promising approach towards evaluating adaptive systems is to decompose the adaptation process and evaluate the system in a "piece-wise" manner. This paper presents a decomposition model that integrates two previous proposals. The main "stages" identified are: (a) collection of input data, (b) interpretation of the collected data, (c) modeling of the current state of the "world", (d) deciding upon adaptation, and (e) applying adaptation.
User Control over User Adaptation: A Case Study BIBAFull-Text 443-447
  Xiaoyan Peng; Daniel L. Silver
The A theory of user expectation of system interaction is introduced in the context of User Adapted Interfaces. The usability of an intelligent email client that learns to filter spam emails is tested under three variants of adaptation: no user modeling, user modeling with fixed (optimal) spam cut-offs, and user modeling with user adjustable spam cut-offs. The results supported our hypothesis that user control over adaptation is preferred because the user can maintain the system's interaction state within a region of user expectation. This remains true even when performance of the system (accuracy of spam filtering) degrades because of errors in user control (adjustment of spam cut-offs).
Towards User Modeling Meta-ontology BIBAFull-Text 448-452
  Michael Yudelson; Tatiana Gavrilova; Peter Brusilovsky
The paper proposes meta-ontology of the user modeling field. Ontology is meant to structure the state-of-the-art in the field and serve as a central reference point and as a tool to index systems, papers and learning media. Such ontology is beneficial for both the user modeling research community and the students as it creates a shared conceptualization of the known approaches to building user models and their implementations.

Web Site Navigation Support

Evaluation of a System for Personalized Summarization of Web Contents BIBAFull-Text 453-462
  Alberto Díaz; Pablo Gervás; Antonio García
Existing Web personalized information systems typically send to the users the title and the first lines of the chosen items, and links to the full text. This is, in most cases, insufficient for a user to detect if the item is relevant or not. An interesting approach is to replace the first sentences by a personalized summary extracted according to a user profile that represents the information needs of the user. On the other side, it is crucial to measure how much information is lost during the summarization process, and how this information loss may affect the ability of the user to judge the relevance of a given document. The system-oriented evaluation developed in this paper indicates that personalized summaries perform better than generic summaries in terms of identifying documents that satisfy user preferences. We also considered a user-centred qualitative evaluation indicating a high level of user satisfaction with the summarization method described, in consonance with the quantitative results.
Social Navigation Support Through Annotation-Based Group Modeling BIBAFull-Text 463-472
  Rosta Farzan; Peter Brusilovsky
Closed corpus AH systems demonstrate what is possible to achieve with adaptive hypermedia technologies. However, they are impractical for dealing with the large volume of open corpus resources. Our Knowledge Sea project explores social navigation support, an approach for providing open corpus personalized guidance that is based on past learners' interaction with the system. The most recent stage of our project focuses on using annotations for social navigation support. We present here Knowledge Sea II, which implements annotation-based social navigation support, and report the results of several classroom studies, which have evaluated this technology.
Discovering Stages in Web Navigation BIBAFull-Text 473-482
  V. Hollink; M. van Someren; S. ten Hagen
Users of web sites often do not know exactly what they are looking for or what the site has to offer. During navigation they use the information found so far to formulate their information needs and refine their search. In these cases users need to pass through a series of pages before they can use the information that will eventually answer their question. Recommender systems aimed at leading users to target pages directly do not provide optimal assistance to these users. In this paper we propose a method to automatically divide web navigation into a number of stages. A recommender can use these stages to recommend pages which do not only match the topic of a user's search, but also the current stage of the navigation process. As these recommendations are more tailored toward the user's current situation, they can provide better assistance than recommendations made by traditional recommender systems.
The Impact of Link Suggestions on User Navigation and User Perception BIBAFull-Text 483-492
  Ion Juvina; Eelco Herder
The study reported in this paper explores the effects of providing web users with link suggestions that are relevant to their tasks. Results indicate that link suggestions were positively received. Furthermore, users perceived sites with link suggestions as more usable and themselves as less disoriented. The average task execution time was significantly lower than in the control condition and users appeared to navigate in a more structured manner. Unexpectedly, men took more advantage from link suggestions than women.

Doctoral Consortium Papers

Modeling Emotions from Non-verbal Behaviour in an Affective Tutoring System BIBAFull-Text 493-495
  Samuel Alexander
Emotions are an important issue in user modeling. This paper presents a proposal for an Affective Tutoring System (ATS) that can recognise emotions through automated facial expression and gesture analysis, and show emotions through an animated agent. The domain of the system will be addition for 8 to 9 year olds. An observational study of human tutors has been conducted as a basis for developing the tutoring strategies of the ATS.
Ubiquitous User Modeling in Recommender Systems BIBAFull-Text 496-498
  Shlomo Berkovsky
The existing personalization services usually base on proprietary and partial user models. This work attempts at evolving inference-based mediation mechanism that will facilitate integrating user models coming from different sources, such as repositories of other service providers and user's personal devices. This will allow obtaining more information about the users and providing more accurate personalization. The efficiency of the above approach will be demonstrated using the techniques from Recommender Systems domain.
User Modelling to Support User Customization BIBAFull-Text 499-501
  Andrea Bunt
The following describes ongoing doctoral research on creating a mixed-initiative framework to help users customize complex interfaces. The framework relies on a rich user model to provide customization suggestions with the goal of improving user performance while maintaining a high level of user satisfaction.
ETAPP: A Collaboration Framework That Copes with Uncertainty Regarding Team Members BIBAFull-Text 502-505
  Christian Guttmann
The organized nature of human collaboration is often used as a metaphor for computational theories of collaboration. Knowledge of collaborators' capabilities and reliability of decision making processes are important factors in collaborative activities. In this thesis, we investigate these factors in the context of an important collaborative activity -- the assignment of team members to tasks.
Towards Explicit Physical Object Referencing BIBAFull-Text 506-508
  Michael Kruppa
The main goal of the work presented in this paper is to determine an optimal strategy for virtual characters performing judicious combinations of speech, gesture and motion in order to disambiguate references to objects in the physical environment. The work is located in the research area of mobile computing and deals with the combination of mobile and stationary devices.
Adaptive User Interfaces for In-vehicle Devices BIBAFull-Text 509-511
  Talia Lavie
Adaptive user interfaces (AUIs) have become the focus of various scientific disciplines and are studied extensively over the last decade. The studies exploring the field investigate a broad range of adaptation methods in different types of applications. Although some progress was made in the study of AUIs, many issues need additional exploring. The objective of this research is to extend previous research on AUI and to examine different levels of adaptivity in AUIs, rather than viewing adaptivity as an all or none process. This research will attempt to identify the levels of adaptivity appropriate for different users, tasks and situations when using AUIs. In particular, the research will assess the effects of different levels of adaptivity on the performance of routine and infrequent tasks. A series of experiments will be conducted to develop and evaluate a model specifying the factors that influence the user's interaction with the AUI. Four different levels of adaptivity will be used, ranging from totally manual to fully adaptive with two intermediate levels. The AUI will be examined in the context of in-vehicle systems. The results of the research are expected to facilitate a better understanding of AUIs, clarify uncertainties and specify the situations in which adaptivity should be beneficial. Finally, the results of this research will assist in-vehicle system designers, by providing guiding principles for designing more usable AUIs.
Agent-Based Ubiquitous User Modeling BIBAFull-Text 512-514
  Andreas Lorenz
The main objective of the thesis is to define and implement a framework for agent-based distributed user-modeling. This paper introduces the approach for applying agent technology and illustrates the research issues in distributing the knowledge about the user among active entities, and distributed user-model acquisition and application methods.
Using Qualitative Modelling Approach to Model Motivational Characteristics of Learners BIBAFull-Text 515-517
  Jutima Methaneethorn
Recent research points to the notion that motivation is a crucial factor when creating Intelligent Learning Environments (ILEs). Yet the research in motivation in tutoring systems has not fully considered relationships between features of ILEs and components of learners' motivational structure. This paper proposes to use a qualitative modelling approach to model motivational characteristics of learners while interacting with an ILE within the context of educational game and narrative.
Improving Explicit Profile Acquisition by Means of Adaptive Natural Language Dialog BIBAFull-Text 518-520
  Rosmary Stegmann
As opposed to implicit user profiling, there are only few explicit approaches, which furthermore suffer from problems that prevent them from being truly viable in practice. In this paper we present an approach to explicit user profiling by means of an adaptive natural language dialog. The dialog adapts to interests the user has mentioned and captures new, not predefined user information, which is stored in a semantically structured explicit user profile.
Modelling User Ability in Computer Games BIBAFull-Text 521-523
  David Storey
User Modelling in computer games is an area that holds much research potential which can lead to practical benefits for computer game players. Our research is looking into the problem of concretely defining what makes a player 'good' at both games in general, specific game genres and individual games. We shall then devise a way of measuring ability to produce numerical rankings. These rankings, after being put through comprehensive evaluation by players, have many potential uses including more in-depth comparison methods, opponent matching and coaching applications.
Constraint-Sensitive Privacy Management for Personalized Web-Based Systems BIBAFull-Text 524-526
  Yang Wang
This research aims at reconciling web personalization with privacy constraints imposed by legal restrictions and by users' privacy preferences. We propose a software product line architecture approach, where our privacy-enabling user modeling architecture can dynamically select personalization methods that satisfy current privacy constraints to provide personalization services. A feasibility study is being carried out with the support of an existing user modeling server and a software architecture based development environment.
Modularized User Modeling in Conversational Recommender Systems BIBAFull-Text 527-529
  Pontus Wärnestål
My research interest lies in investigating user-adaptive interaction in a conversational setting for recommender systems, with particular focus on modularized user model components and the use of a dialogue partner (DP) in such systems.