Context and ContextAware Computing
Omar Khan
CS260, Fall 2006
Background
► Ubiquitous
computing in early 90s:
computing everywhere and “invisible”
► Implication
 Create applications that work seamlessly in
human environments
 Understanding of context
Olivetti Active Badges
► Problem:
locating researchers
► Solution: badge tied to identity, tracked as
researcher moves in building
Assistant sees this view
- knows where researcher is
- can forward call
[Want and Hopper, 1992]
Roadmap
► Understanding
context
► Given an understanding of context, how can
applications use it?
► Example applications
► Two Approaches: Dey and Abowd, Dourish
► Case Study and Discussion
What’s the Context?
Context:
What’s the Context
Context:
Shop in Indonesia
Video
► http://www.youtube.com/watch?v=MTh5nC
N_3K0
Thoughts
► Question:
How do we effectively infer
characteristics of situations and usefully
supplement them with technology?
Goal [Dey and Abowd, 2000]
► Apps
like Active Badges using specific user
context (e.g. location) as application input
► Need representation of context
 Helps to build context-aware applications
►better
embedded in the UbiComp and mobile realms
What is Context?
► By
example
 Location, time, identities of nearby users …
► By
synonym
 Situation, environment, circumstance
► By
dictionary [WordNet]
 the set of facts or circumstances that surround a
situation or event
► Problems:
 New situations don’t fit examples
 How to use in practice?
Operational Definition of Context
“Context is any information that can be used
to characterize the situation of an entity. An
entity is a person, place, or object that is
considered relevant to the interaction
between a user and an application,
including the user and the application
themselves.” [Dey and Abowd, 2000]
Observations
- From point of view of developer
Active Badges
► Application:
help operator forward calls to
researcher at appropriate location
Entity
Researcher
Characteristic Info (context)
Badge ID/Name, location,
?Time of the workday
(morning, lunch, dinner)?
Room
Presence of a phone
Museum Audio Guide Example
► Application:
digital museum guide
Entity
Characteristic Info (context)
Museum
Patron (user)
Museum Audio Guide Example
► Application:
digital museum guide
Entity
Characteristic Info (context)
Museum
Education, age, spoken
Patron (user) language, location in museum,
previously viewed artifacts
Exhibit
What area of museum
Mobile
Interface
battery life
Context Categories
► Recall
Dey’s goal: operational definition for use by
designers and developers
► Once you have entities, want to identify frequently
useful contexts
► Primary Categories
 Answer basic questions like who, what, when, where
 Index into more detailed secondary categories
► Secondary
Categories
 More specific details that may be relevant
Primary Categories
► Identity:
every entity has a unique id
► Location: position, spatial relationships
(latitude/longitude, with friends, near a
Starbucks, in the library)
► Activity: what’s happening in the situation
(touring a museum, reading a book)
► Time: current time, duration of event,
temporal ordering
Secondary Categories
► Indexed
by primary category
► Phone number, address, social network,
etc..
► E.g. identity -> email address, phone
number, etc..
Context-Aware Applications
► “A
system is context-aware if it uses context
to provide relevant information and/or
services to the user, where relevancy
depends on the user’s task.”
Context-Aware Features
1.
Presentation of information and services
•
2.
Automatic execution of services
•
3.
Tour guide, Active Badges
Smart homes (turn off lights, adjust
temperature)
Tagging of context to information for later
retrieval
•
Digital camera meta-data (time, location)
Context Toolkit [Salber et al, 1999]
Active Badges
Discussion
► If
you were designing an application and
you wanted to take advantage of context,
would this framework be helpful?
► Example: cell-phone restaurant locator
► Entities: ?
► Relevant Characteristics (context): ?
► Does this help the designer and user?
Dourish’s View on Context
► What
we talk about when we talk about
context [2004]
► Consider
a central goal of UbiComp –
invisibility of useful technology
 Does not arise from design, but from use and
incorporation into practices [Tolmie et al. 2002]
Dourish’s Context
► Context
is not
 Set of stable features that characterize events
 Representable
► Context
is
 Emergent property of interactions (with people,
objects)
Dourish’s Context
► Previous
approaches to context are
representational: “what is context and
how can it be encoded?”
► Alternative approach uses interactional
model: “how and why, in the course of their
interactions, do people achieve and
maintain a mutual understanding of the
context for their actions”
Implications of Representable
Context
► Context
is:
 Form of information that can be encoded
 Delineable: in advance define what contexts are
relevant for the application
 Stable: determination of relevance of potential
context in an activity can be made once,
reused
 Separable from activity
Context can be encoded
► Alternative:
 You cannot bundle up all the context
 Objects can be contextually relevant
- Dey: relevant info about entities
(people, exhibit, interface, …) is
context
- Dourish: all those things might
be contextually relevant, but they
do not fully describe the context
Context is Delineable
► Alternative:
 Scope of application’s contextual features is
defined dynamically
- Dey: When contexts X, Y, Z
come into play, feature A can be
engaged
- Dourish: problematic
Context is Stable
► Alternative
 Context is an occasioned property
 Particulars of situation and activity matter
- Dey: Relevance of user’s
proximity to an exhibit is always
relevant
- Dourish: highly dependent on
the current situation
Context is separable from Activity
► Alternative
 Context is produced,
maintained and enacted
while doing the activity
 Dey sort of agrees, but
for him activity is very
general
Practice
► Practice:
find meaning in the world by
seeing what actions we can engage in
 Computer scientist example
► Context
concerns:
 How actions become meaningful in certain
situations
 Practice
► Practice
evolves => Context Evolves
Implications on Design
► Technology
becomes meaningful as
individuals engage with it
► Use may not align with designer’s
conception:
 Unexpected uses (e.g. SMS)
 Generally used features “particularized”
differently (e.g. our different uses of folder
hierarchies)
Implications on Design
► Predefined
contexts will likely fail
► “How can ubiquitous computing support the
process by which context is continually
manifest, defined, negotiated, and shared?”
► Support evolution of meaning through
practice
Example Application
► Structures
in Information Spaces
► User places items in a two dimensional
space, interact directly with data
► System suggests relationships, user may
Suggested by
work off those suggestions
Application
Discussion
► Building
an application
► Current applications
Building an Application
► If
you were designing an application and
you wanted to take advantage of context,
would these frameworks help?
► Example: cell-phone restaurant locator
► Do these help the designer and user?
Dey
Entities: ?
Relevant Characteristics
(context): ?
Context-Aware App?
Dourish
Context Engaging/Producing
App?
Both?
Case Study: Web Apps
► Do
they match up with our discussion of context?
► How effective are they?
► What are the problems?
► What can they learn from the views of context
discussed here?
► Dey: Context can be represented and processed
► Dourish: Context is emergent. Applications
should help users produce new meanings and
contexts
Gmail
Yahoo! ZoneTag
del.icio.us