Migration of Legacy systems into Object Oriented Architecture

1
Community-Driven Adaptation
Iqbal Mohomed, Alvin Chin, Jim Cai, Eyal de
Lara Department of Computer Science University of
Toronto
Results
User Study
Mobile devices, such as cell phones and
network-enabled personal digital assistants
(PDAs) are fast becoming the platforms of choice
for accessing Web content. Unfortunately, most
Web content requires more resources than those
available on mobile devices, and therefore needs
to be adapted to operate properly in the mobile
environment. Community-Driven Adaptation (CDA)
is a novel approach to automatic content
adaptation that adapts content based on the
datas relevance to the task the user is
performing, and leverages the aggregate
experiences of large numbers of users accessing
common content.

Experimental Setup
Motivation

• Tasks
• Users complete task for each web application
  (car show, e-store and online map)
• Users increase fidelity of images if they need
  to (indicated by right image below)

• High device heterogeneity with different and
  limited resources (memory, bandwidth, resolution,
  processor speed)
• Impractical to create content specifically for
  all mobile devices
• Want to automatically transform content
  on-the-fly based on device and user
  characteristics
• State of the art
• Automatically make adaptation decision based on
  rules or constrains
• Automatic solver finds optimized solution
  meeting all constrains and maximizes user
  satisfaction
• Adaptation decisions based on small sets of
  constrains on broad sets of content
• Does not take into account content semantics and
  datas relevance to the users task

Wastage sent by various adaptation policies
Car show application. Users go through an online
car show and record license plates for each car,
if it has one.
Extra clicks required by various adaptation
policies
Community-Driven Adaptation

• Users grouped into communities based on common
  characteristics
• Members of same community have common
  requirements for adaptation
• Adapted content suitable for some is likely to
  be acceptable to others
• Relies on feedback from previous users to drive
  adaptation process
• Predictive adaptation service learns how to
  adapt content by observing how users alter
  previously adapted content to make it more useful
  to them

E-Store application. Users enter an online
e-commerce store and are asked to buy a PDA with
a red mark at the bottom right, a cellular phone
with a label that reads "Sierra Wireless, and an
Aibo robot wearing a blue and silver necklace.
E-store Fidelity class distribution for users
Architectural Framework

Average fidelity and size for images for all 3
web sites
Online map application. Users are asked to find a
path between the main library and a subway
station from the U of T campus map.
Research supported by Bell University Labs Alvin
Chin holds Walter C. Sumner Fellowship