Recent Changes - Search:

Home

Course Info

2012 Class Notes

2012 Students

Student Groups

..........................

2009 Students

Student Groups

..........................

2006 Class

2006 Faculty

2006 Evaluations

Lecture Notes

2006 Students

Student HWs

API Groups

Student Groups

Opportunities

Wiki Docs

edit SideBar

PersonalKnowledgeBase

Group members

Google group

Forecasting matrix 1

The idea of this matrix is to list the interdependencies of our product variables.

  • 0 = independent
  • 1 = dependent
 storageencodinginterfaceknowledge modelarchitectureprivacy controls
storage111111
encoding111111
interface001000
knowledge model010111
architecture101011
privacy controls111011
cultural acceptance011101
learning styles001101
semantic web111111
data transfer rates111010
political/legal environment011011
data acquisition110111
tags000110

Constraints

* = essential
consistency across ontologies
knowledge expression*
knowledge storage*
knowledge retrieval*
i/o*

Components

* = essential
knowledge model*
mobile access and geo tagging*
cloud storage*
personal wiki*
visual language or interface
multiple views of knowledge*

Forecasting matrix 2

This second matrix reflects the adoption of ideas from the semantic web.

Creativity techniques

techniqueresulting ideas
replacement or substitution 

Presentation

Micro API specification

Draft 1

FunctionParametersDefaults
listen()List of applications or contacts that initiate recording of knowledgeMobile contacts
read()List of applications or contacts to be scanned for knowledgeCalendar, mail client, mail contacts, editors, blogging tools, social media accounts
passwords()List of application passwordsCreates universal password that when used gives PKB access to applications in list
store()Personal cloud configurationDirects knowledge to free, hosted triplestore with limited space
access()List of applications with access to PKBCalendar
languages()List of languages PKB can expect to findEnglish
expiration()Different password which can be shared and becomes valid after a period of inactivityInactivity period is 1 year; data is deleted after that

Use cases

PKB for medical professionals

Use caseMedKB
history: created 2012-04-22 by Dana Scott
DescriptionProvide doctors and other medical professionals with an unobtrusive way of interacting with knowledge specific to their field on a continuous basis by recording, organizing, and storing facts derived from conversations with patients and staff, procedures, diagnoses, prescriptions, etc.
sources:
Actors1. Doctors primary
2. Nurses secondary
3. Hospital administration secondary
4. Patients secondary
5. Laboratory workers secondary
6. Systems provided by hospitals and drug/device manufacturers secondary
Assumptions1. The ML/NLP aspects of the technology must be able to perform with high accuracy in the biological domain.
2. The system must be unobtrusive.
3. Doctors must accept and trust the software before assigning it sensitive tasks.
4. The software must reduce administrative overhead for the user and allow him or her to focus on tasks that require their expertise.
Steps1. A doctor's MedKB checks the hospital's records for the current day and retrieves all relevant knowledge on patients, procedures, treatments, drugs, devices, etc.
2. MedKB apprises the doctor, appropriate staff and systems of possible constraints such as resource scheduling, drug interactions, etc. that might affect the goals on the itinerary.
3. Through the course of his or her work, the doctor updates MedKB with new observations, insights, facts, orders, etc.
4. MedKB routes this new information to people and systems that require or can benefit from this knowledge.
5. Over longer durations such as months, years, and even careers, unique cases can be identified and reviewed via MedKB for continuous improvement of the individual doctor as well as methods that involve teams of specialists. This knowledge can also be repurposed into case studies to train future medical professionals.
Issues1. Accuracy of ML/NLP algorithms including classification and speech recognition.
2. Security of confidential patient knowledge captured by the system.

Final presentation

Edit - History - Print - Recent Changes - Search
Page last modified on May 07, 2012, at 04:18 PM