Profile

Cover photo
dawn ahukanna
Lives in UK
5,148 followers|400,121 views
AboutPostsCollectionsPhotos

Stream

dawn ahukanna

Shared publicly  - 
 
Looking to make a start on your Android app?
 
We're happy to announce the new Study Jams website available to the Google Developer Groups community. This website is a valuable resource to get information about Android for Beginners and an easy way for participants to sign up for the Udacity course: http://developerstudyjams.com/
Enter in short description
6 comments on original post
1
Add a comment...

dawn ahukanna

Shared publicly  - 
 
Health and sports goodness from +DAREBEE.
 
Being Smart

For those who have little time to train but still want to control their body and enjoy full health our new understanding of fascia, the connective tissue that acts as a single organism throughout the body could have been the perfect metaphor for the social media revolution we are experiencing :) 
13 comments on original post
2
Add a comment...

dawn ahukanna

Shared publicly  - 
 
The brain is an organic container of dynamic wiring, with electricity generating chemical triggers controlled by hormones.
"Something" is bound to go awry.
 
Error in the Network Effect

Privately I have always thought that mental illness stems from cross-wiring in the brain going a little awry. Now we have some evidence that the dearth of connectivity in the brain may be the cause of schizophrenia.
A new study provides researchers with their first biological handle on the disorder, and helps explain why it often begins at a relatively young age.
2 comments on original post
1
Add a comment...

dawn ahukanna

Shared publicly  - 
 
Content strategy via +Adrian Warman
 
4 Dimensions of Enterprise Content Strategy

I've been working out a chart lately that is helping me cut through miscommunication and confusion in my conversations about content strategy. I'm finding this chart useful because I talk with people in many pockets of the CS space, and it sometimes takes me a while to unbury my own and others' assumptions about which part of the space we're talking about (or what the space even comprises).

So far, the people I've shared this chart with are also finding it useful. I share its current form here—probably just a snapshot in its evolution—in the hopes that some of you will also find it useful and will help me make it even more useful (or maybe point out that this has already been done, in which case there's no need for a new chart).

Caveat: I'd like to resist the temptation to load lots more stuff onto the visual. (There's so much to say! So much to clarify! Witness the new CS glossary that this community is launching.) I add new elements with caution because every addition sacrifices a bit of simplicity—sometimes worth doing but always a tradeoff.

My goal: To capture the essentials that define the enterprise CS space. Why? So that all of us talking about CS can quickly identify which part we're talking about (and whether it's a tiny part or the whole shebang). Why? So that our conversations can be more productive. Why? Because productive. End of whys.

HT: +Robert Rose, +Joe Pulizzi, +Rahel Anne Bailie, +Michele Linn, +Hilary Marsh, +Michael Priestley, +Scott Abel, +Ann Rockley, +Kristina Halvorson, +Karen McGrane, +Michael Andrews, +Destry Wion, +Andrea L. Ames, +Meghan Casey, +Margot Bloomstein, +Melissa Breker, +Noz Urbina, +Rachel Lovinger, +Jonathon Colman, +Paula Land, +Kevin Nichols, +Tom Johnson, and others whose ideas have contributed. 

The chart you see below is the original version that I posted here Jan. 22, 2016. Please see the most recent version:
https://drive.google.com/folderview?id=0BwT3Gz_wWBmTeDA1em5MWEotQWc&usp=sharing
37 comments on original post
1
Add a comment...

dawn ahukanna

Shared publicly  - 
 
What did you learn today?
Beyoncé has american football match as warm up act!
That's different and novel, no? {>_<}
 
How To Invest In Yourself. — Life Learning
6 comments on original post
1
Add a comment...

dawn ahukanna

Shared publicly  - 
 
The onion decomposition of a network

The onion decomposition of a network is a recently introduced network fingerprinting technique. It can be used to identify structural properties of a network, such as ”tree-like” structure, ”loopy” structure, or the property of having an ”interesting“ sub-network.

Examples of networks that can be analysed in this way include purely mathematical networks; physical networks like transportation networks and power grids; and social networks such as collaboration graphs. The picture shows three mathematical networks: (a) the Cayley tree, in which all but the terminal nodes are connected to exactly three other nodes; (b) the Erdős–Rényi random graph; and (c) the square lattice, in which each node not on the boundary of the network is connected to exactly four others.

Under the picture of each network is its onion spectrum. This can be thought of as a summary of the onion decomposition of the network, which we describe below. The number and nature of the pieces of the onion spectrum can be used to gain insight into the structure of the original network.

The onion decomposition was defined a few months ago in the paper Network structure at multiple scales via a new network statistic: the onion decomposition by Laurent Hébert-Dufresne, Joshua A. Grochow, and Antoine Allard (http://arxiv.org/abs/1510.08542), and the picture shown here comes from that paper. The decomposition is a refinement of the previously known k-core decomposition, and it is called the onion decomposition because it is reminiscent of peeling an onion.

To explain how this works, we use concepts from graph theory. A graph is a mathematical object that can be used to model a network. A graph consists of a set of nodes, called vertices, and a set of edges, which connect pairs of vertices. We do not allow our graphs to contain more than one edge connecting the same pair of vertices, or to have any edges that connect a vertex to itself. The degree of a vertex is the number of edges emanating from it.

To decompose a graph, one first removes all vertices of degree zero (isolated points); these vertices form the 0-shell of the graph. To form the 1-shell, one removes all vertices of degree 1, followed by all the vertices that become vertices of degree 1 as a result of this process, then iterating the procedure again and again until there are no more vertices of degree 0 or 1. The vertices that are removed at each iteration of the construction of the 1-shell are called the layers of the 1-shell. The construction of the 2-shell (and higher shells) follows a similar procedure: at each iteration, one removes all vertices of degree 2 or lower, until this is no longer possible.

The onion spectrum shown in the picture shows the relative abundance of vertices in each layer within each shell of the corresponding graph. Each connected piece of the picture is shown in a different colour and corresponds to a shell, and each dot corresponds to each successive layer of that shell. The vertical scale is logarithmic, which means that a straight line corresponds to exponential behaviour. The meaning of the colours in the pictures of the actual networks is different: the size of a vertex corresponds to its degree, and the colour of a vertex shows whether it is shallow or deep in the network. The original k-core decomposition gives less information, because it only keeps track of the k-shells, and not the layers within each shell.

The paper shows how to use the onion spectrum to read off properties of the original network. Exponential decay in the spectrum, like in the Cayley tree case, is indicative of the graph having a tree-like structure, whereas sub-exponential decay, like in the random graph case, suggests that the graph has a lot of circuits. It is a nice exercise to show that the Cayley tree consists of one big 1-shell, and the square lattice consists of one big 2-shell; it follows from this that the onion spectrum of each graph has only one connected component. 

Networks like these last two, whose onion spectrum has fewer components than a random graph, tend to be disassortative, which means that nodes tend to be connected to nodes that are somehow ”unlike” themselves. One way to illustrate this is to colour the vertices of the square lattice in alternating colours, like a chess board; something analogous can be done for the Cayley tree. 

Conversely, a network with more components than a random graph will tend to be assortative, meaning that nodes tend to be connected to nodes that are somehow like themselves. A good example of this that is studied in the paper is the collaboration graph of condensed matter physics papers on the arXiv preprint server. Some other real world graphs that are studied are the graph of a power grid, which is tree-like and assortative, and the network of roads in Pennsylvania. The road network graph has one particular shell that exhibits a change in its decay rate, which is indicative of the network having an interesting sub-network. In fact, the road system consists of several essentially isolated components.

Relevant links

Here's another post by me about assortative and disassortative networks: https://plus.google.com/101584889282878921052/posts/cHo5dMTQdsW
That post links to two other relevant posts by me, whose links I provide below for convenience.

A post by me about the mathematics of social networks: https://plus.google.com/101584889282878921052/posts/YV7j9LRqKsX

A post by me about the random graph: https://plus.google.com/101584889282878921052/posts/34guwy4ftWX

Wikipedia has more information about k-cores on its page about degeneracy in graph theory: https://en.wikipedia.org/wiki/Degeneracy_(graph_theory)

I'm glad that Google has fixed the discouraging bug that was corrupting the plus-one counts of posts. On the negative side, it looks like the selectedpapers.net site (which was associated with the spnetwork hashtag) is no more. Can anyone confirm or deny this?

#mathematics #sciencesunday  
22 comments on original post
1
Add a comment...
Have her in circles
5,148 people
Kimberly Maracucci's profile photo
Mustansar Iqbal's profile photo
Mrjn Ml's profile photo
Kelly Linda's profile photo
Anna Morgan's profile photo
IRoo Case's profile photo
Emrul Islam's profile photo
Mark Horwitz's profile photo
Vadilson Malaquias dos Santos's profile photo

dawn ahukanna

Shared publicly  - 
 
Do you IA?
 
Information architecture supports all aspects of the web experience. It enhances accessibility, and reinforces the efficacy and authenticity of sites. Yet, Abby Covert argues that IA is still an elusive concept, with a vast contingent of those who practice it groping at best, and copying obsolete strategies at worst. Only a fearless commitment to talking about IA—including the failures, the confusion, and the Eureka! moments—will bring this essential element out of the shadows.  

http://alistapart.com/article/pain-with-no-name
Information architecture is the problem…and the solution.
View original post
1
Add a comment...

dawn ahukanna

Shared publicly  - 
 
More Icons.
 
Introducing icons 2.2 – our latest update covers everything from sentiments to subways, and expands our system icon library to 933. Visit design.google.com/icons to see all 41 new looks.
9 comments on original post
2
David Lawhorn's profile photo
 
Thanks for sharing.
Add a comment...

dawn ahukanna

Shared publicly  - 
 
This is why Software as a Service (SaaS) fails to gain users trust.
At least if the user runs the software on their machine, they decide when to decommission the software and storage.

Via +Ade Oshineye.
 
It's getting time, GPhotos reaches parity with Flickr in usability and features, and time to get the hell out of that service. Which is a fucking shame, the last iteration had so much potential.
Yahoo announced yesterday that it will be cutting 15% of its workforce, or about 1,700 jobs, in an effort to bring the flagging Internet company back into
View original post
1
Add a comment...

dawn ahukanna

Shared publicly  - 
 
"a concept called 'institutionally constrained technology adoption'.
In other words, even though a more effective solution was possible and easily available, deploying it would increase the risk of undermining the ruling authorities.

This is at the heart of the technology trap posed by 'eating our own dog food'. It is an extremely good way of building product improvement into everyday processes, but only if the strengths and weaknesses of the product are honestly identified, admitted, and addressed. In other words, an organization that espouses the dog food principle must be willing to acknowledge criticism as well as compliments.
And that is impossible not easy for some organizations to do.
 
Technology traps: The danger of &#39;eating your own dog food&#39;.
The always excellent Bruce Schneier refers to an interesting paper talking about the debate between longbow and crossbow as a weapon system, and how the choice was affected - perhaps catastrophically - by national politics and stability. We'll return to tha...
View original post
1
Add a comment...

dawn ahukanna

Shared publicly  - 
 
🎶take good care of yourself🎶 - the 3 degrees.
 
Time to ditch the self-doubt and hold your head up high.
2 comments on original post
1
Add a comment...

dawn ahukanna

Shared publicly  - 
 
Pretty.
 
Dark side & bright side - love the shadow movement ;)

Bees & Bombs creation

#math   #processing   #animation   #circles  
1 comment on original post
4
Add a comment...
dawn's Collections
People
Have her in circles
5,148 people
Kimberly Maracucci's profile photo
Mustansar Iqbal's profile photo
Mrjn Ml's profile photo
Kelly Linda's profile photo
Anna Morgan's profile photo
IRoo Case's profile photo
Emrul Islam's profile photo
Mark Horwitz's profile photo
Vadilson Malaquias dos Santos's profile photo
Collections dawn is following
Basic Information
Gender
Female
Story
Tagline
Software Alchemist - Turning base code into precious applications. Devsigner == 'Dev'eveloper + De'signer'
Introduction
All views and opinions expressed by me are my own only and not those of my employer, any other organisation or individual.
Places
Map of the places this user has livedMap of the places this user has livedMap of the places this user has lived
Currently
UK
Links