| Introduction to the special issue on statistical and probabilistic methods for user modeling | | BIB | Full-Text | 1-4 | |
| David Albrecht; Ingrid Zukerman | |||
| A hybrid approach for improving predictive accuracy of collaborative filtering algorithms | | BIBAK | Full-Text | 5-40 | |
| George Lekakos; George M. Giaglis | |||
| Recommender systems represent a class of personalized systems that aim at
predicting a user's interest on information items available in the application
domain, operating upon user-driven ratings on items and/or item features. One
of the most widely used recommendation methods is collaborative filtering that
exploits the assumption that users who have agreed in the past in their ratings
on observed items will eventually agree in the future. Despite the success of
recommendation methods and collaborative filtering in particular, in real-world
applications they suffer from the insufficient number of available ratings,
which significantly affects the accuracy of prediction. In this paper, we
propose recommendation approaches that follow the collaborative filtering
reasoning and utilize the notion of lifestyle as an effective user
characteristic that can group consumers in terms of their behavior as indicated
in consumer behavior and marketing theory. Emanating from a basic
lifestyle-based recommendation algorithm we incrementally proceed to the
development of hybrid recommendation approaches that address certain dimensions
of the sparsity problem and empirically evaluate them providing further
evidence of their effectiveness. Keywords: Recommender systems; Collaborative filtering; Personalization; Lifestyle | |||
| Efficient and non-parametric reasoning over user preferences | | BIBAK | Full-Text | 41-69 | |
| Carmel Domshlak; Thorsten Joachims | |||
| We consider the problem of modeling and reasoning about statements of
ordinal preferences expressed by a user, such as monadic statement like "X is
good," dyadic statements like "X is better than Y," etc. Such qualitative
statements may be explicitly expressed by the user, or may be inferred from
observable user behavior. This paper presents a novel technique for efficient
reasoning about sets of such preference statements in a semantically rigorous
manner. Specifically, we propose a novel approach for generating an ordinal
utility function from a set of qualitative preference statements, drawing upon
techniques from knowledge representation and machine learning. We provide
theoretical evidence that the new method provides an efficient and expressive
tool for reasoning about ordinal user preferences. Empirical results further
confirm that the new method is effective on real-world data, making it
promising for a wide spectrum of applications that require modeling and
reasoning about user preferences. Keywords: Preference elicitation; Ordinal utility function; Reasoning over
preferences; Support vector machines; Kernels | |||
| Personalizing influence diagrams: applying online learning strategies to dialogue management | | BIBAK | Full-Text | 71-91 | |
| David Maxwell Chickering; Tim Paek | |||
| We consider the problem of adapting the parameters of an influence diagram
in an online fashion for real-time personalization. This problem is important
when we use the influence diagram repeatedly to make decisions and we are
uncertain about its parameters. We describe learning algorithms to solve this
problem. In particular, we show how to modify various explore-versus-exploit
strategies that are known to work well for Markov decision processes to the
more general influence-diagram model. As an illustration, we describe how our
techniques for online personalization allow a voice-enabled browser to adapt to
a particular speaker for spoken dialogue management. We evaluate all the
explore-versus-exploit strategies in this domain. Keywords: Personalization; Influence diagrams; User-model adaptation; Planning;
Dialogue management; Speech recognition | |||
| Improving command and control speech recognition on mobile devices: using predictive user models for language modeling | | BIBAK | Full-Text | 93-117 | |
| Tim Paek; David Maxwell Chickering | |||
| Command and control (C&C) speech recognition allows users to interact
with a system by speaking commands or asking questions restricted to a fixed
grammar containing pre-defined phrases. Whereas C&C interaction has been
commonplace in telephony and accessibility systems for many years, only
recently have mobile devices had the memory and processing capacity to support
client-side speech recognition. Given the personal nature of mobile devices,
statistical models that can predict commands based in part on past user
behavior hold promise for improving C&C recognition accuracy. For example,
if a user calls a spouse at the end of every workday, the language model could
be adapted to weight the spouse more than other contacts during that time. In
this paper, we describe and assess statistical models learned from a large
population of users for predicting the next user command of a commercial
C&C application. We explain how these models were used for language
modeling, and evaluate their performance in terms of task completion. The best
performing model achieved a 26% relative reduction in error rate compared to
the base system. Finally, we investigate the effects of personalization on
performance at different learning rates via online updating of model parameters
based on individual user data. Personalization significantly increased relative
reduction in error rate by an additional 5%. Keywords: Command and control; Language modeling; Speech recognition; Predictive user
models | |||
| Adaptive testing for hierarchical student models | | BIBAK | Full-Text | 119-157 | |
| Eduardo Guzmán; Ricardo Conejo | |||
| This paper presents an approach to student modeling in which knowledge is
represented by means of probability distributions associated to a tree of
concepts. A diagnosis procedure which uses adaptive testing is part of this
approach. Adaptive tests provide well-founded and accurate diagnosis thanks to
the underlying probabilistic theory, i.e., the Item Response Theory. Most
adaptive testing proposals are based on dichotomous models, where he student
answer can only be considered either correct or incorrect. In the work
described here, a polytomous model has been used, i.e., answers can be given
partial credits. Thus, models are more informative and diagnosis is more
efficient. This paper also presents an algorithm for estimating question
characteristic curves, which are necessary in order to apply the Item Response
Theory to a given domain and hence must be inferred before testing begins. Most
prior estimation procedures need huge sets of data. We have modified
preexisting procedures in such a way that data requirements are significantly
reduced. Finally, this paper presents the results of some controlled
evaluations that have been carried out in order to analyze the feasibility and
advantages of this approach. Keywords: ITS; Student diagnosis; Adaptive testing; Item response theory; Statistical
kernel smoothing | |||
| Complementary computing: policies for transferring callers from dialog systems to human receptionists | | BIBAK | Full-Text | 159-182 | |
| Eric Horvitz; Tim Paek | |||
| We describe a study of the use of decision-theoretic policies for optimally
joining human and automated problem-solving efforts. We focus specifically on
the challenge of determining when it is best to transfer callers from an
automated dialog system to human receptionists. We demonstrate the
sensitivities of transfer actions to both the inferred competency of the
spoken-dialog models and the current sensed load on human receptionists. The
policies draw upon probabilistic models constructed via machine learning from
cases that were logged by a call routing service deployed at our organization.
We describe the learning of models that predict outcomes and interaction times
and show how these models can be used to generate expected-utility policies
that identify when it is best to transfer callers to human operators. We
explore the behavior of the policies with simulations constructed from
real-world call data. Keywords: Spoken dialog systems; Machine learning; Human-machine systems;
Probabilistic user modeling; Complementary computing | |||
| Discovering stages in web navigation for problem-oriented navigation support | | BIBAK | Full-Text | 183-214 | |
| Vera Hollink; Maarte van Someren | |||
| Users of web sites often do not know exactly which information they are
looking for nor what the site has to offer. The purpose of their interaction is
not only to fulfill but also to articulate their information needs. In these
cases users need to pass through a series of pages before they can use the
information that will eventually answer their questions. Current systems that
support navigation predict which pages are interesting for the users on the
basis of commonalities in the contents or the usage of the pages. They do not
take into account the order in which the pages must be visited. In this paper
we propose a method to automatically divide the pages of a web site on the
basis of user logs into sets of pages that correspond to navigation stages. The
method searches for an optimal number of stages and assigns each page to a
stage. The stages can be used in combination with the pages' topics to give
better recommendations or to structure or adapt the site. The resulting
navigation structures guide the users step by step through the site providing
pages that do not only match the topic of the user's search, but also the
current stage of the navigation process. Keywords: Navigation support; Information needs; Web usage mining; Navigation stages | |||
| A content-collaborative recommender that exploits WordNet-based user profiles for neighborhood formation | | BIBAK | Full-Text | 217-255 | |
| Marco Degemmis; Pasquale Lops | |||
| Collaborative and content-based filtering are the recommendation techniques
most widely adopted to date. Traditional collaborative approaches compute a
similarity value between the current user and each other user by taking into
account their rating style, that is the set of ratings given on the same items.
Based on the ratings of the most similar users, commonly referred to as
neighbors, collaborative algorithms compute recommendations for the current
user. The problem with this approach is that the similarity value is only
computable if users have common rated items. The main contribution of this work
is a possible solution to overcome this limitation. We propose a new
content-collaborative hybrid recommender which computes similarities between
users relying on their content-based profiles, in which user preferences are
stored, instead of comparing their rating styles. In more detail, user profiles
are clustered to discover current user neighbors. Content-based user profiles
play a key role in the proposed hybrid recommender. Traditional keyword-based
approaches to user profiling are unable to capture the semantics of user
interests. A distinctive feature of our work is the integration of linguistic
knowledge in the process of learning semantic user profiles representing user
interests in a more effective way, compared to classical keyword-based
profiles, due to a sense-based indexing. Semantic profiles are obtained by
integrating machine learning algorithms for text categorization, namely a
naïve Bayes approach and a relevance feedback method, with a word sense
disambiguation strategy based exclusively on the lexical knowledge stored in
the WordNet lexical database. Experiments carried out on a content-based
extension of the EachMovie dataset show an improvement of the accuracy of
sense-based profiles with respect to keyword-based ones, when coping with the
task of classifying movies as interesting (or not) for the current user. An
experimental session has been also performed in order to evaluate the proposed
hybrid recommender system. The results highlight the improvement in the
predictive accuracy of collaborative recommendations obtained by selecting
like-minded users according to user profiles. Keywords: User modeling; Collaborative filtering; Content-based filtering; Hybrid
recommenders; Machine learning; Neighborhood formation in recommender systems;
WordNet | |||
| Adaptive, intelligent presentation of information for the museum visitor in PEACH | | BIBAK | Full-Text | 257-304 | |
| Oliviero Stock; Massimo Zancanaro | |||
| The study of intelligent user interfaces and user modeling and adaptation is
well suited for augmenting educational visits to museums. We have defined a
novel integrated framework for museum visits and claim that such a framework is
essential in such a vast domain that inherently implies complex interactivity.
We found that it requires a significant investment in software and hardware
infrastructure, design and implementation of intelligent interfaces, and a
systematic and iterative evaluation of the design and functionality of user
interfaces, involving actual visitors at every stage. We defined and built a
suite of interactive and user-adaptive technologies for museum visitors, which
was then evaluated at the Buonconsiglio Castle in Trento, Italy: (1) animated
agents that help motivate visitors and focus their attention when necessary,
(2) automatically generated, adaptive video documentaries on mobile devices,
and (3) automatically generated post-visit summaries that reflect the
individual interests of visitors as determined by their behavior and choices
during their visit. These components are supported by underlying user modeling
and inference mechanisms that allow for adaptivity and personalization. Novel
software infrastructure allows for agent connectivity and fusion of multiple
positioning data streams in the museum space. We conducted several experiments,
focusing on various aspects of PEACH. In one, conducted with 110 visitors, we
found evidence that even older users are comfortable interacting with a major
component of the system. Keywords: Adaptive mobile guides; Multimodal user interfaces; Personalized information
presentation; Personal visit report | |||
| The role of human factors in stereotyping behavior and perception of digital library users: a robust clustering approach | | BIBAK | Full-Text | 305-337 | |
| Enrique Frias-Martinez; Sherry Y. Chen | |||
| To deliver effective personalization for digital library users, it is
necessary to identify which human factors are most relevant in determining the
behavior and perception of these users. This paper examines three key human
factors: cognitive styles, levels of expertise and gender differences, and
utilizes three individual clustering techniques: k-means, hierarchical
clustering and fuzzy clustering to understand user behavior and perception.
Moreover, robust clustering, capable of correcting the bias of individual
clustering techniques, is used to obtain a deeper understanding. The robust
clustering approach produced results that highlighted the relevance of
cognitive style for user behavior, i.e., cognitive style dominates and
justifies each of the robust clusters created. We also found that perception
was mainly determined by the level of expertise of a user. We conclude that
robust clustering is an effective technique to analyze user behavior and
perception. Keywords: Digital libraries; Human factors; Stereotypes; Robust clustering;
Perception; Behavior | |||
| Navigation behavior models for link structure optimization | | BIBAK | Full-Text | 339-377 | |
| Vera Hollink; Maarten van Someren | |||
| Analysis of existing methods for automatic optimization of link structures
shows that these methods rely heavily on assumptions about the preferences and
navigation behavior of users. Authors often do not state these assumptions
explicitly and do not evaluate whether the assumptions are consistent with the
actual behavior of the users of the site. This is a serious deficiency as
experiments with simulated users show that incorrect assumptions can easily
lead to inefficient link structures. In this work we present a framework that
gives a systematic overview of alternative assumptions. On the basis of the
framework we can select a set of assumptions that best matches the navigation
behavior of the users in the site's log files. We also present a method for
optimizing hierarchical navigation menus on the basis of the selected
assumptions. This method can be used interactively under full control of a web
master. The system proposes modifications of the structure and explains why
these modifications lead to more efficient menus. Evaluation by means of a case
study shows that the modifications that are proposed effectively reduce the
expected navigation time while preserving the coherence of the menu structure. Keywords: Link structure optimization; Web site efficiency; User behavior models;
Model selection; Navigation menus | |||
| Adaptive feedback generation to support teachers in web-based distance education | | BIBAK | Full-Text | 379-413 | |
| Essam Kosba; Vania Dimitrova; Roger Boyle | |||
| This work examines the application of user-adapted technologies to address
problems experienced in web-based distance education. We have proposed an
approach to support distance learning instructors by offering advice that
points at problems faced by students and suggests possible activities to
address these problems. The paper describes an original feedback generation
framework which utilises student, group and class models derived from tracking
data in web course management systems, and follows a taxonomy of feedback
categories to recognise situations that are brought to the instructors'
attention. The results of an empirical study in an online learning course point
at benefits of the generated feedback to both instructors and students.
Teachers can get a better understanding of their students by knowing what
problems they may be facing, when they are behind or ahead of their peers, who
can help them and how, and what roles can be assigned in discussion forums.
This, in turn, can have a positive effect on students who can receive feedback
tailored to their needs and problems. The evaluation study points at issues
that can be related in general to planning empirical evaluations of
user-adapted systems in realistic web-based learning settings. Keywords: (Semi-)automatic advice generation; Personalised feedback; Teacher
support/help; Empirical evaluation; Intelligent course management systems;
Distance learning | |||
| Modeling the progression of Alzheimer's disease for cognitive assistance in smart homes | | BIBAK | Full-Text | 415-438 | |
| Audrey Serna; Hélène Pigot; Vincent Rialle | |||
| Smart homes provide support to cognitively impaired people (such as those
suffering from Alzheimer's disease) so that they can remain at home in an
autonomous and safe way. Models of this impaired population should benefit the
cognitive assistance's efficiency and responsiveness. This paper presents a way
to model and simulate the progression of dementia of the Alzheimer's type by
evaluating performance in the execution of an activity of daily living (ADL).
This model satisfies three objectives: first, it models an activity of daily
living; second, it simulates the progression of the dementia and the errors
potentially made by people suffering from it, and, finally, it simulates the
support needed by the impaired person. To develop this model, we chose the
ACT-R cognitive architecture, which uses symbolic and subsymbolic
representations. The simulated results of 100 people suffering from Alzheimer's
disease closely resemble the results obtained by 106 people on an occupational
assessment (the Kitchen Task Assessment). Keywords: Cognitive assistance; Cognitive modeling; Error simulation; Cognitive
architecture; Smart home; Alzheimer's disease | |||
| Inferences, suppositions and explanatory extensions in argument interpretation | | BIBAK | Full-Text | 439-474 | |
| Sarah George; Ingrid Zukerman | |||
| We describe a probabilistic approach for the interpretation of user
arguments that integrates three aspects of an interpretation: inferences,
suppositions and explanatory extensions. Inferences fill in information that
connects the propositions in a user's argument, suppositions postulate new
information that is likely believed by the user and is necessary to make sense
of his or her argument, and explanatory extensions postulate information the
user may have implicitly considered when constructing his or her argument. Our
system receives as input an argument entered through a web interface, and
produces an interpretation in terms of its underlying knowledge representation
-- a Bayesian network. Our evaluations show that suppositions and explanatory
extensions are necessary components of interpretations, and that users consider
appropriate the suppositions and explanatory extensions postulated by our
system. Keywords: Discourse interpretation; Suppositions; Explanatory extensions;
Probabilistic approach; Bayesian networks | |||
| Predicting time-sharing in mobile interaction | | BIBAK | Full-Text | 475-510 | |
| Miikka Miettinen; Antti Oulasvirta | |||
| The era of modern personal and ubiquitous computers is beset with the
problem of fragmentation of the user's time between multiple tasks. Several
adaptations have been envisioned that would support the performance of the user
in the dynamically changing contexts in which interactions with mobile devices
take place. This paper assesses the feasibility of sensor-based prediction of
time-sharing, operationalized in terms of the number of glances, the duration
of the longest glance, and the total and average durations of the glances to
the interaction task. The data used for constructing and validating the
predictive models was acquired from a field study (N=28), in which subjects
performing mobile browsing tasks were observed for approximately 1 h in a
variety of environments and situations. The predictive accuracy achieved in
binary classification tasks was about 70% (about 20% above default), and the
most informative sensors were related to the environment and interactions with
the mobile device. Implications to the feasibility of different kinds of
adaptations are discussed. Keywords: Time-sharing; Attention; Multitasking; Interruptions; Mobile interaction;
Mobility; Classification; Predictive models; Bayesian networks | |||