|
Wednesday, May 14, 2003
|
|
Time
|
Project
|
Members
|
Description
|
|
1:00-1:30
|
ArchMiner
|
Leah Zagreus
Stephanie Hornung
|
Our team is developing a new exploratory data analysis tool
for the Center for the Built Environment (CBE), a research
group at the College of Environmental Design (http://www.cbe.berkeley.edu).
One of CBE's research projects, the Occupant Satisfaction
Survey, uses an online survey to collect data regarding
occupant satisfaction in relation to a variety of metrics
(such as air quality, acoustics and lighting) in various
buildings. As a supplement to their current reporting
capabilities, ArchMiner is intended to enable dynamic
exploration of the data collected by the survey. By
combining the familiar cross-tab format with the charts
used in the reporting tool, we hope users will be able to
visually compare question responses within as well as
across datasets. This will allow them to more easily
investigate hypotheses and observe trends in building
occupant satisfaction with respect to various building
technologies.
|
|
1:30-2:00
|
Asset Management at Pacific Gas & Electric Company
|
Dieyana Ruzgani
|
This project spans the design and implementation of an asset
management system to handle the activation and maintenance on
PG&E's wireless telephony assets. Currently the process is
manual and paper-based, with the involvement of PG&E's help
desk. Our new system moves users to an automated,
database-driven website solution for the Utility's 22,000
staff.
|
|
2:00-2:30
|
CampusConnect
|
Beth Manning
Yin Lau
Christina Heyl
Catherine Lai
|
Many activities exist that require more than one person. It is
often difficult to find like-minded partners of similar ability
levels. Current systems are often highly specialized or not
geographically oriented, which makes the matching process less
efficient and effective. CampusConnect will be a system that
facilitates matching people with similar interests and skills.
|
|
2:30-3:00
|
Veterinary Information System
|
David Rieger
|
This web-based patient-tracking system has been created to report,
triage and provide care to patients. Veterinary technicians report
health-conditions patients present under their care and health
staff triage and follow-up on those reports. Health staff have
the ability to longitudinally track problems to facilitate problem
solving and help fine-tune health care delivery. This system has
been implemented on the UC Berkeley campus and has been nominated
by Veterinary staff to be presented at a conference focused on
animal medicine.
|
|
Break
|
|
3:20-3:50
|
Re-employment Information System
|
Derek Jaeger
Shannon Ye
Guoping Li
|
Our project seeks to establish a framework for a prototypical
retiree re-employment model and assess its relevance to future
planning. Our primary goal is to provide management with a
reliable and cost-effective method for identifying and contacting
clusters of retiree candidates suitable for staffing existing
openings.
|
|
3:50-4:20
|
The Berkeley XML Application Infrastructure
|
Patrick Garvey
Marc Gratacos
John Jairo Leon
Calvin Smith
|
Our project consists of a set of tools, methodologies, and
artifacts that compose an infrastructure for conducting model
driven software development using XML Schema as a basis. Our
project demonstrates how Information Architecture, one of the
core competencies of a SIMS education, can be used to generate
deployable systems with a minimum of code customization. Systems
built in this manner will be more robust, flexible, and easy to
maintain. We applied Document Engineering methods and developed
tools that embodied best practices in UI design techniques. We
collaborated with campus IT and operational units to ensure that
our tools and techniques are relevant and usable for real-world
applications as well as for course projects. Our collaboration
produced a reference implementation of the UC Berkeley Course
Approval System. We intend that this implementation will become
the model for further application development here at SIMS and
at Berkeley.
|
|
4:20-4:50
|
Group Voice
|
Diane Ghorbani
Leticia Valdez
John Fritch
|
Decision-making usually involves obtaining the opinions of
group members, but when the group is unable to meet at a
central location, opinion gathering becomes challenging for
leaders. Many groups try to manage this process via email or
electronic bulletin boards, but these both present problems.
Opinions can be overlooked and summarization is time consuming.
In addition, the opinions and documents used in making the
decision are not stored in a central location for reference.
Group Voice addresses all of these problems via a centralized
web application that enables an organization to: present
questions; collect opinions; provide leaders with a summary
of the opinions; and archive questions, documents and
opinions for future reference.
|
|
Thursday, May 15, 2003
|
|
Time
|
Project
|
Members
|
Description
|
|
1:00-1:30
|
GSO Executive Dashboard
|
Vishal Badiani
Naidu Buyyala
Emily Liggett
Holly Liu
|
The GSO Executive Dashboard system was developed to meet the user
requirements and the business needs of the Global Sales Operations
group at Sun Microsystems, Inc. Our final product is a web based
executive dashboard system that provides the following
functionality: an intuitive UI, role based information access
depending, drilldown functionality that lets executives quickly
narrow down and pinpoint drivers causing variation in business
performance, administrative functionality, trend graphs that
allow users to easily understand data significance over time,
a metric data dictionary and an "add comments" feature that
allows executives to attach comments to individual metrics.
The system increases decision-making power, reduces cycle time
by five fold, reduces chances of making errors, increases
availability, reliability and scalability and was developed with
a user centric UI design methodology.
|
|
1:30-2:00
|
Ubiquitous Computing and Privacy in the Workplace
|
Osbaldo Cantu
|
Hospitals are some of the first locations and nurses the first
employees to experience and interact with real world ubiquitous
computing applications. The deployment of ubiquitous computing
in the workplace raises many privacy questions. Technical systems
like the nurse locator badge system installed in hospitals possess
many of the properties of ubiquitous computing. To better
comprehend the privacy concerns and experiences of nurses using
the locator badge system, we collected information from nurses
through online discussion groups, interviews, focus groups, the
badges themselves (technology), the policies governing data
collection and use of the system, and the actual use of the system.
Information about the experiences and reactions of the nurses were
valuable in shaping our research and understanding of privacy in
this context.
|
|
2:00-2:30
|
Semantic Extraction with Wide-Coverage Lexical Resources
|
Behrang Mohit
|
Portability and domain independence are critical challenges
for Natural Language Processing (NLP) systems. Semantic
Extraction is an NLP task that pertains to the assignment
of semantic bindings to short units of text (usually sentences).
NLP problems such as Information Extraction, Question Answering
Systems and Text Classification Systems should benefit from
Semantic Extraction. I have used two manually-built knowledge
bases (WordNet and FrameNet) to automate Semantic Extraction.
My prototype system shows promising results when compared to
existing algorithms. As part of this work, I compiled a large
and semantically-rich information extraction pattern set and
lexicon, and will make this available to the NLP community in
the near future.
|
|
2:30-3:00
|
Muni Map
(working
prototype)
|
Margaret Law
Kaichi Sung
|
Increasing numbers of public transit riders are using computers
to access information about buses, subways, and other modes of
mass transportation. Today, transit systems worldwide have online
counterparts to their paper map alternatives, many borrowing
common visualization and presentation techniques from popular
mapping websites. The result: a spectrum of approaches to the
seemingly simple problem of getting from point A to point B.
Many fall short of tapping the full potential of the online
environment. This project compares a variety of public transit
sites and explores techniques for solving a set of fairly
standard mapping problems. We then apply the more usable
approaches to a web- and kiosk-based prototype designed for
the San Francisco Municipal Railway (Muni) system. In doing so,
we hope to offer a user-centric model for use by transit systems
everywhere.
|
|
Break
|
|
3:20-3:50
|
iCoach, a Virtual Career Counselor
|
Lin Jiang
Phoebe Liu
|
A successful job interview depends what you say and more
importantly, how you say it. A study done at UCLA a few
years ago revealed that the impact of a performance was
based 7 percent on the words used, 38 percent on voice
quality, and 55 percent on nonverbal communication. We
are building an interactive multimedia system iCoach to
assist job seekers in improving their speaking skills in
a job interview. By streamlining and combining the audio/video
capturing, editing, audio analysis, and performance report
generating processes into one cohesive system users get
instant feedback for improving their speaking skills.
Interacting with a virtual career counselor, a prospective
job hunter will answer a series of questions and have his
or her interaction recorded by a video camera. The system
then performs audio analysis on the recording and generates
a performance report detailing the user's speaking
characteristics, such as energy, tone variation, pitch range
and the length of time the user takes to answer a question.
Based on the performance report, the user can choose to redo
individual questions or the entire session, and visualize the
results in order to compare with different trials.
|
|
3:50-4:20
|
Information Technology for Energy and Maintenance
Management in Buildings
|
Luis Villafana
|
Commercial buildings are often times generators of wasted
resources, in the form of energy and labor. For the past
year I have been involved in the research of resource
reduction through the use of information systems in commercial
buildings. We have determined that resources can be better
managed through the deployment of an integrative information
system that compiles and process information gathered through
a building's computerized maintenance management systems (CMMS),
control systems, and occupants at different levels.
First, by including building occupants in an information system
through an interface, they can be utilized as sensors providing
actual building information and report maintenance problems.
Second, by processing this information through information
retrieval methods, corrective maintenance actions can be
recommended to the maintenance personnel. And third, by
integrating all of the available control system and CMMS
information, the building maintenance process can be automated.
|
|
4:20-4:50
|
Chronkite
|
Grace Jeon
Natalia Perelman
Ethan Eismann
|
Chronkite is an online tool that allows its users to quickly
track the most current news about topics of interest. It
improves news reading by providing means of filtering and
organizing news, news related images and weblogs. Avid online
news readers can easily integrate Chronkite's results into their
News Aggregators.
|
|
Friday, May 16, 2003
|
|
Time
|
Project
|
Members
|
Description
|
|
10:30-11:00
|
CampusConnect
|
Beth Manning
Yin Lau
Christina Heyl
Catherine Lai
|
Many activities exist that require more than one person. It is
often difficult to find like-minded partners of similar ability
levels. Current systems are often highly specialized or not
geographically oriented, which makes the matching process less
efficient and effective. CampusConnect will be a system that
facilitates matching people with similar interests and skills.
|
|
11:00-11:30
|
The Berkeley XML Application Infrastructure
|
Patrick Garvey
Marc Gratacos
John Jairo Leon
Calvin Smith
|
Our project consists of a set of tools, methodologies, and
artifacts that compose an infrastructure for conducting model
driven software development using XML Schema as a basis. Our
project demonstrates how Information Architecture, one of the
core competencies of a SIMS education, can be used to generate
deployable systems with a minimum of code customization. Systems
built in this manner will be more robust, flexible, and easy to
maintain. We applied Document Engineering methods and developed
tools that embodied best practices in UI design techniques. We
collaborated with campus IT and operational units to ensure that
our tools and techniques are relevant and usable for real-world
applications as well as for course projects. Our collaboration
produced a reference implementation of the UC Berkeley Course
Approval System. We intend that this implementation will become
the model for further application development here at SIMS and
at Berkeley.
|
|
11:30-12:00
|
GSO Executive Dashboard
|
Vishal Badiani
Naidu Buyyala
Emily Liggett
Holly Liu
|
The GSO Executive Dashboard system was developed to meet the user
requirements and the business needs of the Global Sales Operations
group at Sun Microsystems, Inc. Our final product is a web based
executive dashboard system that provides the following
functionality: an intuitive UI, role based information access
depending, drilldown functionality that lets executives quickly
narrow down and pinpoint drivers causing variation in business
performance, administrative functionality, trend graphs that
allow users to easily understand data significance over time,
a metric data dictionary and an "add comments" feature that
allows executives to attach comments to individual metrics.
The system increases decision-making power, reduces cycle time
by five fold, reduces chances of making errors, increases
availability, reliability and scalability and was developed with
a user centric UI design methodology.
|
|
Break
|
|
12:15-12:45
|
Semantic Extraction with Wide-Coverage Lexical Resources
|
Behrang Mohit
|
Portability and domain independence are critical challenges
for Natural Language Processing (NLP) systems. Semantic
Extraction is an NLP task that pertains to the assignment
of semantic bindings to short units of text (usually sentences).
NLP problems such as Information Extraction, Question Answering
Systems and Text Classification Systems should benefit from
Semantic Extraction. I have used two manually-built knowledge
bases (WordNet and FrameNet) to automate Semantic Extraction.
My prototype system shows promising results when compared to
existing algorithms. As part of this work, I compiled a large
and semantically-rich information extraction pattern set and
lexicon, and will make this available to the NLP community in
the near future.
|
|
12:45-1:15
|
Muni Map
(working
prototype)
|
Margaret Law
Kaichi Sung
|
Increasing numbers of public transit riders are using computers
to access information about buses, subways, and other modes of
mass transportation. Today, transit systems worldwide have online
counterparts to their paper map alternatives, many borrowing
common visualization and presentation techniques from popular
mapping websites. The result: a spectrum of approaches to the
seemingly simple problem of getting from point A to point B.
Many fall short of tapping the full potential of the online
environment. This project compares a variety of public transit
sites and explores techniques for solving a set of fairly
standard mapping problems. We then apply the more usable
approaches to a web- and kiosk-based prototype designed for
the San Francisco Municipal Railway (Muni) system. In doing so,
we hope to offer a user-centric model for use by transit systems
everywhere.
|
