This doesn't seem like you need fancy linguistics ML to fix. You take 90s-era search engine tech and add a database of attributes about every kind of product th

This doesn't seem like you need fancy linguistics ML to fix. You take 90s-era search engine tech and add a database of attributes about every kind of product there is, and take a guess at whether "without" is a search term, a search result modifier, or a filter for an attribute of a product. When you display the results, simply ask the user if that was right; if it wasn't, ask them if they preferred the other filter method. Use those responses as a corpus to train the algorithm. I mean, context is key, right? You're on Amazon and your first search term is "shirts". Unless their is a band called "shirts without stripes", the user wants shirts. The rest of the query is probably some filter of that product. You know shirts sometimes have stripes. It's not a one-size-fits-all algorithm, but it's simple enough that the user should end up with the results they wanted.