Simran-B
commented
3 years ago We don't have a part of speech (POS) tagger to analyze sentences, so the following hacks come to mind:
- Create a stopword list ("in", "near", "with", "a", "an"...) and simply search for the remaining terms. Possibly maintain a field which has all relevant text concatenated together, so that we can find all terms in this single field.
- Use regular expressions to deconstruct common sentences and extract type, place and features. Reminders in Slack work like that.
Query that searches descriptions of rentals.
Ex: Places in Florida with a pool.