Data Science in the NOC and Beyond
Clayton. A. Hollister
May 29, 2018
Introduction
• We are a small friendly company with the same
big challenges of any corporation.
• Although the scale may differ, we both collect
analytic data for our web site, we both have
data centers to manage and so on.
• Like the early telescopes which gave humans a
new perspective of our universe; viewing your
organization through a data lens provides a
new way to understand your business.
What Do We Do
• Our Data Science team combines raw
data from several sources for analysis.
• We convert the data to a matching format
and perform extensive mathematical
testing resulting in a graphical
representation.
Taking this …………………… to this
Real World Data Science
• You know your business and we know data
science. We work with you to imagine the
possible, find the feasible, and develop the
execution plan. In many situations humans find
themselves like a tooth on a gear too close to
the process to see the whole machine.
• A famous example recounts the birth of the
early automobile by saying if Henry Ford had
done a marketing survey and asked people
what they wanted; their answer would have
been a faster horse!
Real World Data Science Example
• One of our customers had problems managing their
volume of trouble tickets which we can file under big data
causes big problems.
• Our team was able to analyze their data and found a
problem with the way they had configured their network
hardware causing a port storm. We found other problems
that their in house team had not found as they had never
seen all of the data in the same format and in one place.
• The customer was able to make a simple configuration
change to remove the port storm issue which improved
their network operations data management efficiency over
30%. Other recommended changes were made resulting
in more improvements.
Web Site Analytics
• Our experience in this area both in house and for
customers has shown it to be very easy to miss important
data due to a localized maximum hiding other data.
• For example when you look at a chart which categorizes
all of your site visitors it will indicate a large number of
visitors from .com sites. This is because in part .com sites
are the most prevalent on the web.
• This will make the other categories look like insignificant
numbers, however as we found on our own web site when
we removed the large outlier .com numbers we realized a
lot of non-profit organizations were interested in a group of
services we offer. This triggered us to change our
marketing campaign.
Raw Web Analytics Data
Humans find it easier to understand their
data pictorially in charts – not like this
Hidden Data
The larger numbers for .com, net and unknown made the
“non-profit” bar seem insignificant in comparison.
Revealed Data
Once the larger values were filtered out the drill down
revealed a large number of downloads were being made by
non-profit organizations in Germany.
Can We Help You?
• If you are tired of searching through raw
data and want meaningful repeatable
graphic reports we can help.
• We would love to help you uncover the
hidden data in your organization. Please
use our contact form at this link.
• https://www.sinenomine.net/contact-us

Data science in the noc and beyond

  • 1.
    Data Science inthe NOC and Beyond Clayton. A. Hollister May 29, 2018
  • 2.
    Introduction • We area small friendly company with the same big challenges of any corporation. • Although the scale may differ, we both collect analytic data for our web site, we both have data centers to manage and so on. • Like the early telescopes which gave humans a new perspective of our universe; viewing your organization through a data lens provides a new way to understand your business.
  • 3.
    What Do WeDo • Our Data Science team combines raw data from several sources for analysis. • We convert the data to a matching format and perform extensive mathematical testing resulting in a graphical representation. Taking this …………………… to this
  • 4.
    Real World DataScience • You know your business and we know data science. We work with you to imagine the possible, find the feasible, and develop the execution plan. In many situations humans find themselves like a tooth on a gear too close to the process to see the whole machine. • A famous example recounts the birth of the early automobile by saying if Henry Ford had done a marketing survey and asked people what they wanted; their answer would have been a faster horse!
  • 5.
    Real World DataScience Example • One of our customers had problems managing their volume of trouble tickets which we can file under big data causes big problems. • Our team was able to analyze their data and found a problem with the way they had configured their network hardware causing a port storm. We found other problems that their in house team had not found as they had never seen all of the data in the same format and in one place. • The customer was able to make a simple configuration change to remove the port storm issue which improved their network operations data management efficiency over 30%. Other recommended changes were made resulting in more improvements.
  • 6.
    Web Site Analytics •Our experience in this area both in house and for customers has shown it to be very easy to miss important data due to a localized maximum hiding other data. • For example when you look at a chart which categorizes all of your site visitors it will indicate a large number of visitors from .com sites. This is because in part .com sites are the most prevalent on the web. • This will make the other categories look like insignificant numbers, however as we found on our own web site when we removed the large outlier .com numbers we realized a lot of non-profit organizations were interested in a group of services we offer. This triggered us to change our marketing campaign.
  • 7.
    Raw Web AnalyticsData Humans find it easier to understand their data pictorially in charts – not like this
  • 8.
    Hidden Data The largernumbers for .com, net and unknown made the “non-profit” bar seem insignificant in comparison.
  • 9.
    Revealed Data Once thelarger values were filtered out the drill down revealed a large number of downloads were being made by non-profit organizations in Germany.
  • 10.
    Can We HelpYou? • If you are tired of searching through raw data and want meaningful repeatable graphic reports we can help. • We would love to help you uncover the hidden data in your organization. Please use our contact form at this link. • https://www.sinenomine.net/contact-us