4 ms·

I've never been lucky(?) enough to reach a scale or level of domain messiness where the standard method of indexing things naievely in elasticsearch or even typ

• ckorhonen 3y ago
  The main failure mode here I find is nuances in more specialized datasets. Elasticsearch is great, powerful and easy to work with once you get the hang of it, but is very open-ended with lots of room to build a sub-optimal solution. Fuzzy matching and boosting often trip people up or lead to folks shooting themselves in the foot relevancy-wise. If you want really great results, you need to spend the time crafting your query, dealing with synonyms ("pop" vs. "soda"), stemming, typos, negative boosts ("non-alcoholic", in the example) etc. It's not necessarily hard, just often forgotten or not included in the initial scope.
• blacksmith_tb 3y ago
  Many if not most ecom sites will be using a vendor for search (e.g. Bloomreach) so they're sending a feed of product data to their vendor, just a subset of their product data (which may not have enough metadata for some facets people might reasonably expect). That means there's lag, and also that the integration is stiff and brittle, hard to change without time and money. Even then, whether they're rolling their own search or using a 3rd party, nine times out of ten their search will be ONLY for products, good luck if you wanted to search for a phrase from their privacy policy; for that or any sort of other non-product content you're forced to use Google/Bing/ddg etc.