=Paper=
{{Paper
|id=Vol-266/paper-1
|storemode=property
|title=Modeling tasks: a requirements analysis based on attention support services
|pdfUrl=https://ceur-ws.org/Vol-266/paper01.pdf
|volume=Vol-266
|dblpUrl=https://dblp.org/rec/conf/jcdl/LaukkanenRM07
}}
==Modeling tasks: a requirements analysis based on attention support services==
https://ceur-ws.org/Vol-266/paper01.pdf
Modeling tasks: a requirements analysis based on
attention support services
Joona Laukkanen Claudia Roda Inge Molenaar
American University of Paris American University of Paris Ontdeknet
147 rue de Grenelle 147 rue de Grenelle Wibautstraat 4
75007 Paris, France 75007 Paris, France 1090 GE Amsterdam
+33 1 40 62 06 82 +33 1 40 62 06 50 +31 20 525 1339
joona.laukkanen@aup.fr croda@ac.aup.fr inge@ontdeknet.nl
ABSTRACT Modeling user tasks in a manner that is both complete and
The task within which a resource is used is a very important operational is far from being an easy undertaking. Based on the
element for the definition of Contextualised Attention Metadata. work done in the Atgentive project [2, 16, 17, 19], this paper
In this paper we discuss the requirements of a task model that discusses how tasks may be modeled in order to support the
allows representing current and potential attention allocation of implementation of attention management services. In the process
the user. And we discuss how such model has been implemented we will also highlight another important relation between CAM
in the AtGentive system. and task modeling, i.e. the fact that not only (as mentioned above)
tasks may be associated to resource access, but also resources
Categories and Subject Descriptors may be associated to task descriptions.
H.3 Information storage and retrieval, H.5.2 User Interfaces, In the context of the Atgentive system a task represents the target
H.1.2 User/Machine Systems, H.1.1 Systems and Information of an attentional focus (e.g. writing a paper, accessing some
Theory: value of information, I.6.5 Model Development, resource, ...). Since we aim at applicability in combination with a
number of different types of applications, the key design issue
General Terms with the definition of tasks has been to make it as application
independent as possible. In particular, the questions of task
Design granularity, task structure, and task attributes, have been
addressed.
Keywords
In section 2 we give a brief description of the Atgentive system.
attention aware systems, attention metadata, contextualized We summarize the goals of the project, introduce the different
attention metadata, learning, task modeling. modules of the system, and explain how the Reasoning Module
provides functionalities supporting users' attention allocation. The
1. INTRODUCTION analysis of such functionalities has provided us with the most
Task models represent a very important element in the definition critical requirements for the AtGentive task model. Whilst the
of Contextualized Attention Metadata (CAM). As CAM aims at AtGentive System aims at providing many task-oriented services,
tracking resources usage, identifying the specific context in which in this paper we concentrate only on those that support
such usage takes place enables a much better understanding of the interruption management and task switching. In section 3 we
value of each resource [10]. The task within which a resource is discuss the requirements that these services impose on task
used is a very important element of such contextual definition. modeling, section 4 briefly overviews the issues commonly
For example, in order to truly understand resource usage it would encountered in task modeling, and section 5 details the AtGentive
be important to distinguish whether a user accesses a book review task model.
because he is writing a research paper, or because he is preparing
a reading list for a course, or because he is selecting gifts from a 2. THE ATGENTIVE SYSTEM
wedding list. These three types of usages correspond to access to The objective of the AtGentive project is to investigate the use of
the resource book review in the context of different tasks. The artificial agents for supporting the management of the attention of
definition of users' tasks is therefore one of the essential elements young or adult learners in the context of individual and
for the identification of the context of resource usage. collaborative learning environments.
The AtGentive system observes the user's activity and generates
interventions aimed at supporting his/her attentional choices.
Such interventions may either be designed to help users sustaining
their current focus of attention (e.g. help user to find the best way
to complete a task), or they may be designed to shift the user's
attention to a different focus (e.g. communicate important
information that has become available).
The main components of an AtGentive system include: one or
more (1) applications, and (2) user tracking components
providing information about the users activity, both these types of
components communicate with (3) a reasoning module – see The reasoning module is designed as an application independent,
figure 1. Applications, users, and tracking modules inform the general purpose entity capable of generating suggestions about
reasoning module about the state of the user and the environment attention management. Within the AtGentive project the
by generating events. The reasoning module supports the user in reasoning module is being tested in the framework of two
his attentional choices by generating interventions that are then different applications: AtgentSchool, and AtgentNet.
sent to the user. AtgentSchool is an eLearning platform for elementary school
Figure 1 – A simple schema of an AtGentive system aged children, and AtgentNet is a virtual community platform
supporting knowledge exchange in knowledge communities.
3. TASK ORIENTED SERVICES IN
ATTENTION AWARE SYSTEMS
In the context of the AtGentive project we have identified several
task-oriented services aimed at supporting learners and
knowledge workers in environments characterized by frequent
interruptions and multi-tasking. These include: interruption
management, support to task switching, orienteering within
resources (e.g. searching and ranking), and self and community
awareness.
For sake of brevity, in this section we only discuss the first two
services with the aim of detecting the characteristics that a task
model should have in order to enable the implementation of such
services.
Interruption management
Interruption management services are services that may either
Events generated by the application either describe the user automatically select the time and mode of the interventions that
activity (e.g. the user has started working on a certain task) or have been generated , or may provide notification services that
relevant changes in the environment sensed by the application help the user making the decision on when to attend newly
(e.g. new information is available that the user could access). available information.
Events generated by users may describe their preferences (e.g. With respect to interruption management, the task model should
"don't interrupt me when I am working on this task”), or provide a enable reasoning about cost/benefits of interruptions and allow
direct feedback on the reasoning module's interventions. determining the most appropriate time for interruption.
Finally, The tracking devices monitoring the user physical state reasoning about cost/benefits of interruptions
and activity may generate events describing for example the user In order to decide whether to interrupt a user, the system must be
keyboard activity, the level of noise in the room, or the presence able to consider the costs / benefits of the interruption.
(or absence) of the user from the screen.
For example, the system could decide to intervene and suggest
On the basis of these events the reasoning module (which is that the user attends some newly available resource if:
implemented as a multi-agent system) tracks what the users
current focus is, creates a list of possible alternative foci, and [a] the resource is relevant to the task currently in focus or
finally, evaluates those alternative foci and, using interventions, [b] the resource is relevant to an inactive or suspended task
communicates those foci (if any) that seem to be most beneficial with a high priority
to support the user attention.
Note that a resource may be relevant to a task both if it is relevant
While processing events, the reasoning module maintains an to the task or if it is relevant for a sub-task of that task.
optimized list of foci that have been identified as most relevant
Further considerations may intervene if enough knowledge about
for the user. Each focus is composed of a target, a priority, and a
the user tasks is available. For example, in case [b] notification
state. Possible states are: current, inactive, or suspended.
may be delayed if the user is about to complete the current task.
Normally one of the foci is active (this is the user's current focus).
Observatory studies report that returning to long term projects is
Suspended foci are inactive foci that have been previously active.
particularly challenging and makes such tasks potentially more
Inactive foci are those that the reasoning module has evaluated as
vulnerable to the harmful effects of interruptions, compared to
interesting for the user but the user has never activated (e.g. the
more common, shorter tasks, such as writing e-mails [7]. The
focus associated to an email that the user has not yet read). The
expected time that a task on average takes to complete, the
priority is an estimate of how important/urgent the task associated
number of subtasks, and the number of windows and resources
to the focus is for the user. The target of the focus is either a user
that need to be available, could help determining if a task is a
task or a message. A user task is an instance of a generic task for
long term project and hence, interrupting it is more costly than
the specific user in the specific situation (see section 5.1). A
interrupting some shorter task.
message is something that needs to be communicated to the user
without any concrete actions related to it, e.g. some motivational Following the above considerations, in order to reason about
feedback for a learner who has completed an assignment. costs/benefits of interruptions, a task model should allow
identifying:
[REQ 1]The user's current task Restoring task context
[REQ 2]The priorities of the current task and of other When task switches and interruptions are frequent, the activities
(inactive or suspended) tasks required to restore the task context of a resumed task can be
expected to result in a significant increase in cognitive load. A
[REQ 3]The resources that may be relevant to a task
diary study tracking the activity of knowledge workers to
[REQ 4]The advancement state of tasks execution investigate these effects, reported that: (1) participants rated
switching to tasks that were previously interrupted to be
Most appropriate time for interruption significantly more difficult than to others, that (2) the resumed
Several studies have demonstrated that the exact time when an tasks were in fact twice as long as other, more short-term projects
interruption is presented may make a very significant difference and that (3) they required significantly more resources than other
on both how easily the information presented is acquired by the tasks [7]. Automatically providing access to such resources when
user, and on how much disruption it generates in the task being a task is resumed would represent a significant help to users.
interrupted [1, 6]. Providing such service requires that:
In order for the system to determine the most appropriate time for [REQ 6]The task model associates to interrupted tasks
interruption, the task model should support the information describing the resources in use when the
[REQ 5]Description of task hierarchies. task was interrupted.
As noted by Bailey et al. [4] when tasks are organized into Task reminders
hierarchies the task model can be used to infer "breakpoints" i.e. Another problem related to switching tasks is one encountered
times when interruptions are less disruptive for the user. Bailey frequently when a task needs to be performed at a specific
and his colleagues [1, 5] represent tasks as two level hierarchies moment (at an absolute time or in response to some event).
composed of coarse events further split into fine events (for Prospective memory failures, which occur when something
example, a coarse event would be the selection of the email cannot be remembered at the right time, may account for up to
application, which would then be further decomposed in selecting 70% of the memory failures in everyday life [14]. This has been
the email application, typing in the username, and typing in the shown to have a very eminent effect on performance in work and
password). The authors then measure the impact of interruptions learning environments. Also, these memory failures intervene
as they occur at various points within these hierarchies and differently in different age groups.
demonstrate that the best times for interruptions correspond to
coarse breakpoints. The availability of such a hierarchical task Providing services that remind users of important dates and
model enables the system to infer the best time for interruption. In deadlines, or notify them of certain events could be used to
the AtGentive system, when there is a switch in the users current alleviate this threat of prospective memory failures daunting so
task, the magnitude of the break in attention is evaluated on the many activities planned to take place in the future. Further, task
basis of the depth of the task in the task hierarchy. Further, a shift reminders could prove particularly useful to help users remember
to the next subtask can be identified as a low strength break in the tasks that they have suspended earlier as a study has reported that
users attention whilst a jump to a task that is not a child or a in fact over 40% of tasks that have been interrupted, are not
parent of the current task may be interpreted as marking a resumed again [15]. For example if a user suspends a task T1 to
stronger break in the users attention. work on another more urgent task T2, the system could remind
him of the interrupted task T1 once T2 has been completed.
If tasks are organized so that lower level tasks divide a higher
level task into logical sub-steps, the level at which a task switch to In order to provide support with task reminders, it will be
a next subtask happens may be used to infer the magnitude of the necessary to allow:
break in the users attention, possibly with a more accurate value. [REQ 7]Associating to tasks information either about the
A switch at a lower level marks a smaller change in attention than time when the task should be executed, or about the
at a higher level. This would however need the task model to events that should trigger the execution (or resumption)
allow specifying if a task does indeed refine the parent task or if of the task
the parent task exists just to group subtasks, as could be in the
case of a math exercise in a learning environment authored as a Task continuation and prioritization
task hierarchy like [Task T1, “exercise 1”, T1.1, “exercise 1.1”, When there are several tasks that the user is working on in
T1.2, “exercise 1.2”, ..., Task 1.2.n, “3 + 9 / 3 = ” ]. In the latter parallel or there simply are several tasks to choose from, for
case a task switch to a next subtasks would actually mark a example when a task has been completed, it could be beneficial
smaller break in user attention on a higher level task than when for the user if there were services that could take off some of the
the switch to the next subtask happens at the level of the concrete cognitive load that is related to choosing the next activity.
leaf tasks with the actual cognitive work. Especially so when the user might not have much knowledge
about the relevant properties of the different tasks (e.g. how long
Support to task switching a task is expected to last).
Major motivation for services supporting task switching comes On the basis of the task structure it is possible to find some
from the observation that people can only focus on one thing at a potential and logical, yet arguably more or less simple,
time and as several authors have indicated [e.g. 18], switching continuation options for the user. More complex and useful
from a task to another is costly. Services supporting the user with guidance can be achieved by applying some timing strategy in the
task switching operations such as restoring task context, task evaluation, maybe by preferring tasks that may be completed
reminders, and support for task continuation require a before their deadline. If the evaluation also considers the priorities
comprehensive task model.
of different tasks or gets otherwise more sophisticated, the document, submit form. In order to support the user in his
reasoning could be expected to have a real effect on the cognitive attentional choices tasks should be described at a level that
effort required from the user. corresponds better to the user's objectives, (e.g. write a paper,
When a user completes a task, enabling a smooth transition to the complete an exercise). This type of task description has been
next activity may entail restoring the context where the choice to suggested by some researchers [11, 13] and corresponds to the
start the now completed task was made. This could amount to one used in the AtGentive project. In order to achieve this
reminding the user of the task that was suspended when the user objective we require that:
moved to the current task or, reminding him of the current task [REQ 9]The task model should allow different types of
sequence (e.g. the next subtask, the next required task or the applications to define their own tasks and task structure
parent of the task in the hierarchy). [REQ 10]The task model should allow describing tasks at
In both situations the requirement for a hierarchical task structure any level of granularity
([REQ. 5]) is reinforced.
Further, elements that will intervene in the evaluation of valid
Task attributes
Failing to provide contextual information within a task is another
continuations include prioritization (already listed as requirement
pitfall of several task modeling efforts. Contextual information
[REQ. 2]) and timing:
such as relevant resources and users, deadlines, complexity,
[REQ 8]Allow the definition of task deadlines priority, state of advancement, and location of the task in a task
On a more sophisticated level also expected duration of tasks, is hierarchy is something that is clearly needed for many services
required, this is listed below as [REQ 14]. supporting attention management. The inclusion of some of these
attributes is represented by several requirements already listed
above, further task attributes we have identified include:
4.ISSUES IN TASK MODELLING
Diaper quotes Shepherd [20] as saying that “'Task' is seldom [REQ 11]Keywords may be associated to tasks.
defined satisfactorily” and continues suggesting that this might Keywords provide a way to relate tasks to resources (e.g. by
actually never be the case [8]. Some difficulties in defining tasks, keyword matching)
such as the specification of application independent task
[REQ 12]Maximum allowed idle time may be associated to
taxonomies, have been repeatedly encountered and seem
tasks
inherently difficult. Some other issues may be easier to address
but need a comprehensive approach. For example whilst it would The Maximum allowed idle time specifies the time limit within
not be difficult to provide adequate contextual information for which the user is expected to act to avoid being recognized as idle
tasks, this information is often missing from task models. This by tracking devices. This information is used both to identify
section briefly overviews what we consider the main open issues breakpoints and to provide help or solicitations to users who seem
in task modeling. to have difficulties continuing a task.
[REQ 13]Task difficulty levels may be associated to tasks
Task taxonomies
One clear problem when modeling tasks is the difficulty of Indications on the difficulty of a task may help in the evaluation
defining a sufficient taxonomy. It would be useful to classify of the cost/benefits of interruptions, as well in the selection of the
tasks, for example, by type of operation. Finding generic actions help to be provided to users.
or operations independent of application types has however [REQ 14]Expected duration of the task may be associated to
proven very difficult [8]. One of the few generic tasks that Diaper tasks.
& Johnson [9] were able to identify in their work on TAKD (Task
This attribute specifies the average expected time to complete the
Analysis for Knowledge Description) was insert. TAKD is a
task. Combined to the task advancement indication ([REQ. 4]), it
method capable of modeling tasks in a wide range of applications
enables a better evaluation of the best time for interruption.
and within this work insert was found common for a number of
Further, task continuation services may implement strategies in
different objects in different application domains (namely
which, under certain conditions, tasks with certain durations (e.g.
microelectonics, automated office applications, and computer
tasks that can be completed quickly) are preferred over other
programming). Inserting could here mean either inserting text in a
tasks.
word processor or a program editor or alternatively inserting
components on a Printed Circuit Board. Whilst it could be [REQ 15]Actors relevant to the task may be associated to
possible to identify some actions possibly totally independent of tasks
application domains, such as insert, the set of such actions seems Relevant actors could for example include a teacher in the case of
to be simply too small. Whilst Diaper [8] does not see the a learning environment or the creator of a resource when the task
development of task taxonomies as totally impossible, it is is simply to attend some resource. In general actors relevant to
obvious that we are far from having such a tool and probably the tasks will be defined within a social network associated to the
definition of ontologies allowing the integration of several such user model. This information is both useful to evaluate the
taxonomies is the most promising direction of research. relevance of newly available information, and to provide
community awareness services.
Task descriptions
Traditionally tasks have been described at the level of the [REQ 16]Support tasks may be associated to tasks (see
application, i.e. tasks correspond to very specific users' actions section 5)
within a specific application, e.g. create document, attach
Currently we assume that most of these parameters are manually A main task can then be described to consist of a number of finer
entered (e.g. by the user himself, or by a teacher setting up a level tasks. Task T1, Writing a paper, could for example consist
learning sequence - as is done in the AtgentSchool application), in of the more concrete tasks T1.1 (do research) T1.2 (write
the future we expect that the system may be capable of generating abstract), ..., T1.n (discuss future work). The hierarchical
estimates of parameters such as maximum allowed idle time, task organization of main tasks allows for varied granularity when
difficulty, expected duration time, etc. by observing how several defining tasks; nothing forces one to define tasks at a finer level
users act on the task, and by inferring the possible behavior of a so for example writing a paper could in some environments be
specific user. modeled as a single high level task if the task is, perhaps, known
to be already well understood by the target users. In another
Recognizing tasks environment the same task could be represented as one with a
Whilst defining tasks, their structure and resources presents, as number of subtasks (possibly on several levels). Besides allowing
described above, a series of challenges, a further, possibly more granular description of task execution, subtasking can be used to
complex challenge is represented by the automatic recognition of distribute support more accurately where it is needed. This could
tasks. This requires that, on the basis of the observation of user's in fact be one way to author tasks; first identify how the entire
actions, the system is capable of matching actions sequences to task needs to be supported (e.g. support for doing research,
specific tasks. The problem here is that if simple sequences of support for writing the abstract, ...) and divide the task in subtasks
actions are observed (such as typing some characters on the accordingly.
keyboard) the system may not have enough semantic information
In defining task structure we have identified further requirements
to associate the action sequence to a specific task. In fact a very
for the task model these include:
large number of higher-level tasks may be associated to simple
action sequences. Within AtGentive we base task recognition on [REQ 17]The task model may include a requirement level
three possible inputs. First, an application, which has a much for a task
better knowledge of the semantics associated to simple user [REQ 18]The task model may include task ordering
actions may recognize that the user is working at a specific task
and communicate this information to the reasoning module. [REQ 19]The task model may include task visibility
Second, AtGentive may use its knowledge about a small subset of These properties are tightly related to the execution of tasks at a
all possible user tasks that are most likely to be performed by the given moment and are useful to support task continuation (see
user at a given time, and use this information to recognize that a 3.2.3), they are briefly described below.
simple action sequence is actually contributing to a specific task. Task requirement level
Third, the user may explicitly indicate that he is performing a
certain task. Tasks may be defined as optional or required. Required sub-tasks
are necessary (i.e. they must be executed) for the completion of
the parent task. Tasks defined as optional allow the user to skip
5.ATGENTIVE TASK MODEL certain sub-tasks in the execution of a main task. In a learning
The task model implemented in AtGentive's Reasoning Module environment some exercise for example, may be marked as
distinguishes between two different categories of tasks: main optional.
tasks and support tasks. Main tasks are in essence anything the
user may decide to do. Support tasks are aimed at helping the user Task ordering
perform a given main task and manage his attention within that The order in which a task's subtasks need to be performed could
task. either be specified as free for the user to choose, or mandated. In a
learning environment an assignment might for example consist of
Generic Tasks versus User tasks reading a book and then writing a summary about it. Here it
Both main tasks and help tasks represent abstract task properties. would make sense to mark the ordering of the assignments
Whenever main tasks, or help tasks are activated concrete subtasks to be mandated.
instances are created as user tasks. This results in creating a
hierarchy of user tasks corresponding to the hierarchy of the main Note that, if ordered execution is required, optional subtasks can
tasks and support tasks. User tasks instantiate all the properties for still be skipped.
the concrete execution of that task, such as a deadline, Task visibility
progression etc, for one particular user. For example, one may Tasks may either be visible or invisible. Invisible tasks are always
have a main task "prepare lecture" which has abstract properties inner nodes in the task hierarchy and allow describing abstract
such a title, and an average expected duration, and is organized in tasks that, although not executable, are useful to conceptualize a
a hierarchy of sub-(main)-tasks such as "collect resources", certain grouping in sub-tasks. A group of root tasks that are not
"create draft", etc. each having their abstract properties. For each related to each other could for example be grouped under one
user, there would then be a corresponding user task structure to common invisible root task. Invisible tasks could also be useful if
actually execute the task, with for example individual deadlines there is a need for a more complex ordering than what otherwise
for those users. would be allowed by the task model (without adding mundane
tasks that only include selections between subtasks).
Main tasks
Main tasks (and the related user tasks) represent actions the user The task model does not support certain task sequencing
might perform, e.g. write a paper, prepare for a meeting, complete constraints. For example its is not currently possible to specify the
an assignment. These tasks can be formed into hierarchies as requirement that the user completes a certain number of subtasks
pleased as all main tasks could have other main tasks as subtasks. (say for example 2 tasks out of 3).
Support tasks objectives of many of the task models presented in the field of
Support tasks help the user in performing various types of human-computer interaction, some of them may be used to guide
activities that the user might attend at different stages of a tasks future development of our model.
execution. For example, a support task might help a confused user
gaining a better understanding of the task at hand, another support
7.ACKNOWLEDGMENTS
The work described in this paper was partially sponsored by the
task could provide some motivational feedback such as statistical
EC under the FP6 framework project Atgentive IST-4-027529-
information about the users time usage after a task is already
STP. We would like to acknowledge the contribution of all
completed.
project partners.
Support tasks differ from main tasks mainly in two ways. First,
they cannot be organized into hierarchies and they do not have 8.REFERENCES
further support tasks themselves. Although hierarchies of support [1] Adamczyk, P. D. and B. P. Bailey, 2004. If not now, when?
tasks might be a valid concept, we didn't identify a pressing need The effects of interruption at different moments within task
for them and therefore we didn't include the concept in the model. execution. Human Factors in Computing Systems: CHI'04,
This helps us avoiding introducing unnecessary complexity in the New York, ACM Press.
model. Essentially for the same reason we don't consider the
[2] AtGentive (2005-2007). ist-4-027529-stp.
concept of support tasks for support tasks. The other key
difference to main tasks is the classification of support tasks. [3] Azevado, R., Hadwin, A.F. (2005) Scaffolding self-regulated
Support tasks are classified in two ways. First they are classified learning and metacognition – Implications for the design of
based on when they will be relevant to the task that they support. computer-based scaffolds. Instructional Science33: 367–379
This could either be before (pre-task support), during (on-task [4] Bailey, B. P., P. Adamczyk, et al. (2006). "A Framework for
support) or after the task (post-task support). In addition, support Specifying and Monitoring User Tasks." Computers in
tasks are classified by the type of support they provide. Human Behavior 22(4): 709-732.
Based on scaffolding models [3, 12, 21, 22] we have divided [5] Bailey, B. P. and J. A. Konstan (2006). "On the Need for
support tasks into four categories: behavior, cognitive, Attention Aware Systems: Measuring the Effects of
metacognitive and motivational support tasks. Cognitive support Interruption on Task - Performance, Error Rate, and
task have a focus on mental activities of the user. Metacognitive Affective State." Computers in Human Behavior 22(4): 685-
support task are directed at orienting, monitoring and evaluating 708.
activities. Behavioral support task are focused on physical
activities of the user. Motivational support tasks are directed [6] Czerwinski, M., E. Cutrell, and E. Horvitz. Instant
towards feelings of the user [12]. The term scaffolding was messaging: Effects of relevance and time. in HCI 2000 -
introduced by Wood, Bruner, and Ross [22] and it is defined as 14th British HCI group Annual Conference. 2000. University
providing assistance to a student on as-needed basis, fading the of Sunderland: British Computer Society.
assistance as the competence increases. The general idea behind [7] Czerwinski, M., E. Horvitz, et al. (2004). A diary study of
scaffolding is that some of the control within the learning task switching and interruptions. Proceedings of the SIGCHI
environment is temporally transferred from the learner to another conference on Human factors in computing systems, Vienna,
more experienced actor to support the learner to acquire all Austria, ACM Press.
abilities to fully self sustain his learning. The scaffold help [8] Diaper, D. (2002). "Scenarios and task analysis." Interacting
supporting the execution of a task that the student could not have with Computers 14(4): 379-395.
done on its own and it is removed when it is no longer necessary.
Especially in innovative learning arrangement where student are [9] Diaper, D. and Johnson, P., 1989. Task analysis and systems
provided with more control of both learning content and learning analysis for knowledge descriptions: theory and application
procedures scaffolds can support them to deal with this increased in training. In : Long, J., Whitfield, A. (Eds.). Cognitive
responsibility. The task support model allows specifying and Ergonomics and Human-Computer Interaction. Cambridge
selecting the support tasks that assists the learning process of University Press, Cambridge, pp. 191-224.
specific learners based on an assessment of their attentional states. [10] Duval, E. (2005). LearnRank: the Real Quality Measure for
Learning Materials (Tematic Dossier 06 December 2005),
6.CONCLUSIONS AND FUTURE WORK Insight - Oservatory for New Technologies and Education.]
In this paper we have identified the major requirements for a
flexible and operational task model supporting the [11] Gonzalez, V. M. and G. Mark (2004). "Constant, constant,
implementation of attention management services. We have also multi-tasking craziness": managing multiple working
indicated how most of these requirements have been implemented spheres. Proceedings of the SIGCHI conference on Human
within the AtGentive system. We consider the work presented in factors in computing systems, Vienna, Austria, ACM Press.
this paper only a starting point for attention oriented task [12] Hadwin, A., L. Wozney, et al. (2005). "Scaffolding the
modeling and the definitions provided will need to be both Appropriation of Self-regulatory Activity: A Socio-cultural
extended and further detailed. We are currently in the process of Analysis of Changes in Teacher–student Discourse about a
evaluating the performance of the AtGentive system in the two Graduate Research Portfolio." Instructional Science 33(5-6):
pilot environments and we trust that such evaluation will 413-450.
significantly contribute to the further development of the [13] Heath, T., M. Dzbor, et al. (2005). Supporting User Tasks
reasoning module as a whole and of the task model in particular. and Context: Challenges for Semantic Web Research.
While the task model we have presented does not have the same Workshop on End-user Aspects of the Semantic Web
(UserSWeb), European Semantic Web Conference [18] Rubinstein, J. S., D. E. Meyer, et al. (2001). "Executive
(ESWC2005), Heraklion, Crete. Control of Cognitive Processes in Task Switching." Journal
[14] Kvavilashvili, L., D. J. Messer, et al. (2001). "Prospective of Experimental Psychology: Human Perception and
memory in children: The effects of age and task Performance 27(4): 763-797.
interruption." Developmental Psychology 37(3): 418-430. [19] Rudman, P. and M. Zajicek (2006). Autonomous agent as
[15] O'Conaill, B. and D. Frohlich (1995). Timespace in the helper – Helpful or Annoying? IAT 2006 - IEEE/WIC/ACM
Workplace: Dealing with Interruptions. CHI ‘95 Conference International Conference on Intelligent Agent Technology,
Companion, Denver, Colorado, United States, ACM press. Hong Kong.
[16] Roda, C., Ed. (2006). AtGentive (IST-4-027529-STP) [20] Shepherd, A., 1998. HTA as a framework for task analysis.
Deliverable D1.3 - AtGentive conceptual framework and Ergonomics 41 (11), 1537-1552.
application scenarios. [21] Zimmerman, B., Schunk, D. (2001). Theories of self-
[17] Roda, C. and T. Nabeth (2006). The AtGentive project: regulated learning and academic achievement: an overview
Attentive Agents for Collaborative Learners. First European and analysis. In self-regulated learning and academic
Conference on Technology Enhanced Learning EC-TEL'06, achievement (2nd ed.) (pp. 1-37). Mahwah, NJ; Erlbaum.
Crete, Greece, Springer. [22] Wood, D., Bruner, J., & Ross, G. (1976). The role of tutoring
in problem solving. Journal of Child Psychology and
Psychiatry 17, 89-100.