Best Practices for Implementing Full-Text Search in Large Databases ??

[marcellosalass]

Hello everyone,

I am working on a project involving a database with millions of records and I need to implement an efficient full text search feature. I have been reading through the documentation here,, which has been incredibly helpful but I still have a few questions I would love to get insights on:

What are the best indexing strategies to optimize full text search performance in large datasets: ??
How do you handle ranking relevance for results, especially when dealing with diverse types of content: ??
Are there any recommended libraries or frameworks that integrate seamlessly with popular DBs like PostgreSQL or MySQL for full text search: ??

I am also curious about edge cases like handling misspellings or partial matches should these be addressed through preprocessing or are there tools that handle them natively: ??

Looking forward to hearing your thoughts and experiences !!

Thanks in advance !!

With Regards,
Marcelo

[EpsilonAdmin]

Hi @marcellosalass

Actually WPFTS could be not that great for you in case you have millions of records (e.g. posts).
The whole Wordpress could be not good for you either, because to handle such amount of data you will need to use specific optimization and caching algorithms and even clustering.

I would like to propose you to use full-text solutions based on Apache Solr or Elastic Search - those tools are implemented in Java and work way faster than PHP+MySQL solution provided by WPFTS and they are especially good for big data.

Alternatively you can wait until we release a new WPFTS version (in 1-2 months) that can use self-hosted Elastic Search engine to keep full-text index and uses WPFTS data collecting algorithms.

Thank you for the question.