“ A Picture” of Enterprise Architecture
Jan 12, 2015
“A Picture” of
Enterprise Architecture
My Assertion
Areas IT can help in 2020:– Knowledge– Operational Efficiency– Staff– Efficient Process
IT can/should be the differentiator
Color 1
Digital Immigrant to Digital Native
Color 2
Defensive Information Sharingto
Offensive Information Sharing
Color 3
On Premise to Cloud
Color 4
Paper to e-BPML
Color 5
Enterprise to Consumer
Color 6
Search to Latent Semantic Analysis
Math is getting better at machine learning (extract knowledge from data)
))_((_*))_(((
))_((_*))_(((
))_((*))_(((
),(
22
11
bnan
ba
ba
ba
inDocWordCountinDocwordCount
inDocWordCountinDocwordCount
inDocwordCountinDocwordCount
DocDocsim
1980s: Pair wise document similarity (document clustering)
1990s: Latent Semantic Analysis (what does the word mean?)
documents1 2 3 4 5 6 7
1 2 3 4 5 6 7 8 9terms
really just matrix multiplication:term vector (query) x strength matrix = doc vector
Paper (n): The full record that supports that claim should be available for detailed examination and critique
We were taught to share these discoveries by publishing a paper or filing a patent after all
the work was done.
Corporate email communicationsGreen = InternalYellow = External
Meanwhile we communicate electronically ...
to networks of people...
… not the objects
What if we,curate the
conversation...
Is the openness of the next generation going to change the scientific process by allowing computers to mine the “human
conversation”?
Scientific Method
Proving a hypothesis.
Knowledge Discovery
Finding the unknown correlations.
Sentiment Analysis for stock market prediction
Brokerage houses are using computers to “micro-trade” stocks based on sentiment analysis of blog sphere.
Figure from Glance, Hurst, Nigam, Siegler, Stockton, & Tomokiyo, KDD’05
My talk from 2 years ago
200% return
Dow Jones
The more we share the smarter it gets.
Googledocs determined the commonality of what I was entering and automatically
completed the rest of the list.
GoogleDocs searched the web.
Next generation versions will be able to tell you what correlations
exist between two seemingly unrelated attributes
We like …
We don’t like …
A picture
Digital Native e-BPML
Latent Semantic AnalysisCloud
Offensive Information Sharing
Consumer
My Assertion
Areas IT can help in 2020:– Knowledge– Operational Efficiency– Staff– Efficient Process
IT can/should be the differentiator
What is your picture?
What colors do you use?