Detecting user's intent from the query

Folks, I am new to the elasticsearch world, but i have been around enterprise search for a while and one feature that I am very used to is the ability to detect the user's intent through some regular matching pattern and shape the specifics of the search accordingly. A common example would be to detect terms in the query that suggest the user is looking for help (query contains "doesn't work" or "can't do X" or "having a problem with Y", etc) and give a small boost to documents containing "troubleshooting category or something like ""troubleshooting" in the title, as opposed to, say specs documents, that would in turn be boosted when the query contains references to some elements of specs, for instance.

I have not run across any discussion on how something like this can be done in Elasticsearch thus far. There is plenty of discussion on how to boost, but not any on how to make a search rule contingent on what is in the query. Can anyone help me with that? Will I need to add a custom code to add query parsing functionality into my search to be able to handle that?

Thanks in advance

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.