Suppose I want to search for any tweets containing “discriminate” or “discrimination” or “discriminated” or “discriminating.” Can I use the stem “discrim” as a keyword and expect to match tweets with any of those word forms or does my query have to list them all ORd together? The discussion in the docs talks about tokenization as it is applied to the keywords, but IDK about the tweets.
No, no substring matching, and text is UTF-8 NFC normalized, (cafe will also match café) with special characters and punctuation (| _ : . Etc) stripped, so the best way is to string together a query with OR like (discrimination OR discriminate OR discriminated)
