On my Answers experience, I’m finding there is a conflict between taxonomy and name.
I have a healthcare account, and so when I search for a healthcare professional’s name on Answers, predictive search populates the name as a suggestion. However, when clicking on the name, results based on a taxonomy that shares the same characters as the professional’s middle name shows up, and the professional does not show up until the second Results page.
We have run into this before for physician searches using “MD”, so we removed “MD” as a synonym/taxonomy for Muscular Dystrophy, but I do not want to continue removing terms to account for inaccuracies in search results.
What do we suggest for this to not compromise between taxonomy and name searches?
Hi Emily,
After looking at the account, it appears that there is a conflict since the name field of the professionals includes their titles (“MD”, “PNC”, etc.) which correspond to specialties. Since the specialty/category are searchable fields set to NLP Filter, the algorithm looks to be prioritizing those results. There are a couple of options you could try to get more desirable results:
-
Using the Experience Training for NLP Filters (reject unwanted results like the ones you mentioned) to help train the algorithm
-
Try updating your searchable fields in your Answers backend. If the autosuggest populates the entire name field, try setting Name to NLP filter. Or try using Text Search for category/specialty fields (a lot of the searchable fields are currently set to NLP Filter, currently, in the account which could be causing these conflicts)