Similarity Measures and Graph Adjacency with Sets

<p>In my last installment (<a href="https://medium.com/p/648a2f20d389" rel="noopener">Part I</a>), I introduced you to a bit about the process of analyzing an archaeological site with data science. I talked about the frustratingly complex nature of &ldquo;Old Things in Space&rdquo; and how the network of artifacts and locations constitute a bipartite graph.</p> <p>As we worked through the process of creating a two-mode graph and projecting it into its constituent one-mode graphs, however, we ended on something of a cliff-hanger&hellip;&nbsp;<em>everything</em>&nbsp;ended up connected to&nbsp;<em>everything else</em>!</p> <p>Well, sort of. I have to confess that I intentionally lead you down a primrose path straight into a cul-de-sac. All part of my cunning plan, of course, to get to the much more interesting topic of thinking about just what it means for two things to be &ldquo;similar&rdquo; and similarity metrics for&nbsp;<em>sets.</em></p> <p><a href="https://towardsdatascience.com/similarity-measures-and-graph-adjacency-with-sets-a33d16e527e1"><strong>Website</strong></a></p>