Information Technology Reference
In-Depth Information
Shop On-
Line
AND
AND
AND
pre
AND
Check Out
Browse
Items
t 1
AND
t 11
Login
AND
AND
Logout
AND
AND
t 16
Confirm
Check Out
pre
Review
Order
t 2
Use
Cart
Learn About
Items
AND
View and Post
Comments
t 12
AND
pre
pre
AND
AND
pre
pre
t 8
Use
External
Web Site
pre
pre
Provide
Address Info
AND
View Items
pre
AND
Provide
CC Info
t 15
AND
AND
AND
OR
t 10
View
Prices
t 3
t 4
pre
OR
View
Comments
Consult
Printed
Catalog
t 13
t 7
AND
View Basic
Product Info
Provide New
Address Info
AND
pre
t 9
t 14
Use Stored
Address Info
t 5
Add
Comment
Change
Ordering
pre
View Product
Image
t 6
Fig. 1. A goal model
model shows alternative ways by which an on-line store can be used for browsing and
purchasing products.
The graph consists of goals and tasks . Goals - the ovals in the figure - are states of
affairs or conditions that one or more actors of interest would like to achieve [6]. Tasks
- the hexagonal elements - describe particular low-level activity that the actors perform
in order to fulfill their goals. To ease our presentation, next to each task shape a circular
annotation containing a literal of the form t i has been added, which we will use in the
rest of the paper to concisely refer to the task. For example, t 7 refers to the task View
Basic Product Info .
Tasks can be classified into two different categories depending on what the system
involvement is during their performance. Thus, human-agent tasks are to be performed
by the user alone without the support or other involvement of the system under consid-
eration - an external system outside the scope of the analysis may be used though. For
example Consult Printed Catalog ( t 3 ) belongs to this category because it is performed
without involvement of the system. On the other hand, mixed-agent tasks are tasks that
are performed in collaboration with the system under consideration. Thus Add Com-
ment is a mixed-agent task as the user will add the comment and the system will offer
the facility to do so. Another example of a mixed-agent task is View Image : the system
needs to display an image and the user must view it in order for the task to be consid-
ered performed. All tasks of Figure 1 are mixed-agent except for t 3 and t 8 which are
human-agent tasks.
Goals and tasks are connected with each other using AND- and OR-decomposition
links, meaning, respectively, that all (resp. one) of the subgoals of the decomposition
need(s) to be satisfied for the parent goal to be considered satisfied. In addition, chil-
dren of AND-decompositions can be designated as optional . This is visually repre-
sented through a small circular decoration on top of the optional goal. In the presence
of optional goals, the definition of an AND-decomposition is refined to exclude op-
tional sub-goals from the sub-goals that must necessarily be met in order for the parent
goal to be satisfied. For example, for the goal View Items to be fulfilled, the task View
Basic Product Info is only mandatory - tasks View Prices , Change Ordering and View
Product Image may or may not be chosen to be performed by the user.
 
Search WWH ::




Custom Search