Hello - 

How are the candidate pairs for the merge game generated?

For example, when I fired up the 'Merge Items' on the Distributed Game, it asked me if the following two items are the same topic or different:

Nicola Kirsch [Q1986126]
 
Nicky Kirsch [Q13104330]

I don't think they're matches, but how did it pick these two to ask about in the first place?

It sounds like there is a batch process that runs and identifies potential pairs. If so:

- How often does that process run?
- What criteria does it use to predict potential matches?
- Is the list of potential pairs stored somewhere and viewable without clicking through the game?
- Do any of the other matching/merging tools use the same dataset?

Thanks!

-Erik