Since I started my Ph.D about 2 years ago, I have been thinking a lot about where knowledge is contained. This is because one of my objectives during my Ph.D is to advance human knowledge. My research sub-field in nuclear physics is nuclear structure. The holy grail of nuclear structure research (as far as I see it) is to represent the nuclear chart in such a way such that nuclei with similar quantum structure are grouped. The electronic quantum structure has been classified in this way already, we have the periodic table of elements. However, nuclear structure is objectively more complicated and it's unlikely that a simple 2-dimensional table will be an accurate representation. Thus, a large amount of study of nuclear structure involves identifying systematic trends between nuclei. I can add knowledge to the field of nuclear structure by connecting nuclei by a trend / behaviour. Turns out my thoughts on this topic align very closely with the relatively modern (2005) theory of learning called connectivism.

## Connecting nuclear physics and blogging

The posts on this blog are like nuclei. The ideas they contain might be related to others by a topic. Much like how all the nuclei mid shell are well described by the rotor model. I could tag all those nuclei with rotor just like I tag blog posts with a topic. Topics may be connected by posts. The simplest nuclear physics analogy I can think of: K isomers are connected to rotors by nuclei in the Hafnium region. Now this post connects blogging to physics and I tricked you into learning some nuclear physics.

## The knowledge contained in this blog

I wanted a way for me to keep track of the big picture that this blog is presenting to the internet for self preservation and improvement. There's no way a tag cloud was going to be good enough for me. I want to know how topics are connected as that is some representation of the knowledge I have on the blog. I want a network of tags as nodes and posts as links, so I made it. Here it is:

I have bounded the above to a tiny box so that it's viewable on all devices but may look a bit crowded. A more luxurious version can be found on my tags page provided your screen is >600px wide. The graph was made with JavaScript library D3.js and I generated the input data from my blog using Jekyll Liquid.

Nodes are sized based on how many posts have that tag. Nodes with more connections are forced closer together. You can mouse over the nodes to see the posts associated with the topic.

### Generating the data with Jekyll Liquid

This blog is hosted GitHub and compiled automatically when I push to the repository using Jekyll. I wrote a bit of liquid that iterates over tags and posts and generates the input data.

Liquid code: GitHub → Jekyll output: This blog

### Generating the graph

This is the combination of three D3.js examples by: Mike Bostock, d3noob, and Tom Roth.

JavaScript code: GitHub

### Embedding it in your blog

Save the code above in a folder /js on your webhost and with a little bit of styling:

.node {
stroke: #000;
stroke-width: 1px;
}

fill: none;
stroke: #000;
}
.svg-container {
display: inline-block;
position: relative;
width: 100%;
vertical-align: top;
overflow: hidden;
}
svg.svg-content-responsive {
display: block;
margin: auto;
position: relative;
left: 0;
}


And a little HTML:

<div name="tagnet" class="svg-container">
<svg width="600" height="300" class="svg-content-responsive"></svg>
<script src="https://d3js.org/d3.v4.min.js"></script>
<script src="/js/tag-network.js"></script>
</div>


You can embed the tag network anywhere on your Jekyll blog.

## Conclusions

I have been thinking about connectivism as a means of representing knowledge, and encoding information. I have implemented a tag network for this blog to test this idea. The tag network represents the topic connectivity of this blog. At a glance you can see what topics are covered by this blog, how much information about that topic exists, and how to extend your learning to other topics using this blog.

As the author of this blog I can use the emergence of clusters in the tag network as an indicator of concepts that I am covering strongly. I can also use it as a tool to make me cognisant of the knowledge I contain and publish publicly. Unlike in nuclear physics where the universe defines the connections between nuclei, I have control over myself and can use the tag network for self improvement.

I am impressed by how information rich the tag network is and have intentions of applying it in my research of nuclear physics.