search user query pre-processing to search in specific field in solr - django

Question

i am new to solr and django,i am working on implimenting search on a party hall venue search website though i have not worked on the website part just implimenting solr for the search i have indexed party hall venue information in solr with following fields
<field name="id" type="int" indexed="true" stored="true" required="true" multiValued="false" />
<field name="title" type="text_general" indexed="true" stored="true" required="true" />
<field name="slug" type="text_general" indexed="true" stored="true" required="true" />
<field name="description" type="text_general" indexed="true" stored="true"/>
<field name="location" type="text_general" indexed="true" stored="true"/>
<field name="city" type="text_general" indexed="true" stored="true"/>
<field name="area" type="text_general" indexed="true" stored="true"/>
<field name="featured" type="boolean" indexed="true" stored="true" />
<field name="facilities" type="text_general" indexed="true" stored="true" multiValued="true" />
<field name="type_of_venue" type="text_general" indexed="true" stored="true" multiValued="true" />...
there are many other fields which are used only for display of data on results page but only these fields are used to query data
on my website i have a search bar where user can enter any search term and then i search it against title,description,location,facilities
now i did enough reading on how to break down the search terms entered by user to identify which field to find it in so that i can have different template view to show the found matches but couldn't find any technique that will work with solr.
please can anyone suggest me any pre search text processing techniques to make it simpler to generate query to search in solr
thanks in advance

Answer 1

If you index your data carefully (read on examples) there is an easy (but not the best!) way to do this.
Lets say here are your party hall documents
1) party hall - "abc party hall", location - "san jose"
2) party hall - "xyz party hall", location - "san francisco"
3) party hall - "pqr party hall", location - "paris"
4) party hall - "best party hall", location - "san jose"
Lets say your user types "best party hall in san jose" in search bar, ideally you should return #4, 1, right?
You can certainly pre-process your query (complex NLP) to extract potential location data to be used for location field in your query.
For a moment lets take brute-force method here, lets use boolean query and search your full query with all important fields as-is
party_hall: "best party hall in san jose" AND location: "best party hall in san jose"
If you've indexed your data properly (as given in example documents above) you will get best results as expected.
party hall query will not have "san jose" so it will consider document with "best party hall", similarly location field will filter document with "san jose" so technically you should get your best matched documents #4,1. You can use "OR" instead of "AND" but you will get more matched documents (but rank sorting will still be accurate and expected)
Try it with your use case and see if it helps!
p.s - this will work if you use any tokenizer based analyzer like StandardAnalyzer (will not work for KeywordAnalyzer)