Elasticsearch on ChalkBucket

DON'T LURK... Join The Discussion!

Members see FEWER ads

JBS

ChalkBucket Founder
Staff member
Gold Membership
Coach
Proud Parent
Elasticsearch is once again installed on ChalkBucket (hopefully to stay depending on cost). The search results should now be drastically better. Right now we have "stop words" turned on... this means that the following words are excluded from search...

a, an, and, are, as, at, be, but, by, for, if, in, into, is, it, no, not, of, on, or, such, that, the, their, then, there, these, they, this, to, was, will, with

"Word stemming" is also turned on.

Word stemming can improve searching by allowing multiple forms of a word to match. For example, a search for "test" would equally match "tests", "testing", "tested", and so on.

We are also enabling "accent removal and character simplification".

If enabled, before indexing, accents will be removed and other complex character representations converted to simpler ones. This can improve search results by allowing multiple accent variations (esta, está) of a word to match. However, it can also make words match unexpectedly when they differ only by accents.

We do not have any "weighting" of content or dates so all content on ChalkBucket is searched equally.

You are now also able to use Elasticsearch "Simple Query String Syntax"

The simple_query_string supports the following special characters:
  • + signifies AND operation
  • | signifies OR operation
  • - negates a single token
  • " wraps a number of tokens to signify a phrase for searching

Please let me know what questions you have and I will try to answer them.
 
Seems that we will be able to keep our new search software. When we were using AWS (Amazon) we were paying over $500/month for this same service... so we had to shut it off after just over a month. Our new server that is running it has only cost us $6.11 since January 9th. This is for over 35,000 searches that have been performed in that time.
 
Just wanted to bump this thread again. We had some issue with our other server and had to change to a new Elasticsearch server... now we are actually on the Elastic Cloud. I'm waiting to see how cost comes in... hopefully it will be fine.

There is one change from what is stated in the first post though. Since we are trying to "drive discussion" here... and we "lock" threads after a few months... "recency-weighted relevance searches" has been enabled. A half life of 365 days has been set... this means...

If enabled, the search relevance algorithm will be weighted towards newer results based on the half-life value. A document submitted today will be twice as relevant as an identical one submitted a half-life ago.

This thread has been "unlocked" so ask questions if you would like.
 
As you all use the search feature on ChalkBucket... let me know if you have any issue with it.
 

DON'T LURK... Join The Discussion!

Members see FEWER ads

Gymnaverse :: Recent Activity

College Gym News

New Posts

Back