Flow/Architecture/Search/Relevance

ElasticSearch has a default implementation to calculate "relevance" of a "item" compared to your search terms. The way that works is described in detail on their website. This is used for sorting: by default, the search results are ordered by how relevant they appear to be for your search terms.

For Flow, there's support for multiple ways of sorting the data:

  • per relevance (the dafault ES relevance)
  • per topic creation date (ascending & descending)
  • per topic last change date (ascending & descending)

Apart from just using the relevance for sorting, we could also use it for other purposes. That relevance is just a floating point number, e.g.: 0.076713204. I don't think it makes much sense to display that number to users.

If we want to somehow display some other form of "relevance", we could calculate different numbers. E.g. (based on existing mockups) we could calculate the number of keyword matches per topic, and show that instead.