[GSoC Weekly] Using Cassovary to analyze Wikipedia (week 1)

[Szymon Matejczyk]

Hi guys,

I'm Szymon Matejczyk (@szymonmatejczyk) and I'll be working on Cassovary as GSoC student this summer. 

The goal of my project is to use Cassovary to analyze Wikipedia pages graph and use it for entity resolution of short texts.

Last week I've been working on few minor changes needed to load Wikipedia dump to Cassovary:

1. https://github.com/twitter/cassovary/pull/69 -- allow renumbering of Longs or numbering of String nodes so that they can be used in Cassovary
   
2. https://github.com/twitter/cassovary/pull/71 -- reading graphs that have Long or String nodes to Cassovary (using previous PR)
   
3. https://github.com/twitter/cassovary/pull/72 -- improving parallel reading of graphs to move to Futures

Until the end of this week I'm planning to load Wikipedia dumps to Cassovary and do some minor improvements:

1. Test my triangle counting approximated algorithm: https://github.com/twitter/cassovary/pull/62
2. Allow generic node types in Cassovary (for now only Ints are possible) and benchmark performance decrease
3. Enhance traversals

If you have any questions, suggestions, feel free to ask,

Szymon