I studied the results with debugQuery, and I understand how the search is working. The two scores for the two terms are added together, so specifying a boost less than one still adds to the score. For instance, in the first result, content:notes has a score of 1.4892359 and genre:Citation^0.01 has a score of 0.0045761107. Those two numbers are added together to get the total score.
I tried putting a negative boost number in, but that isn't legal. Is there a way to express "genre does not equal Citation" as an optional parameter that I can boost? On Mon, Oct 31, 2011 at 2:26 PM, Paul <p...@nines.org> wrote: > I had been experimenting with bq. > > I switched to boost like you suggested, and get the following error > from solr: "can not use FieldCache on multivalued field: genre" > > But that sounds like the solution I'd want, if it worked, since it's > more flexible than having to reindex. > > On Mon, Oct 31, 2011 at 10:41 AM, Erik Hatcher <erik.hatc...@gmail.com> wrote: >> Paul - look at debugQuery=true output to see why scores end up the way they >> do. Use the explainOther to hone in on a specific document to get it's >> explanation. The math'll tell you why it's working the way it is. It's >> more than just likely that some other scoring factors are overweighting >> things. >> >> Also, now that I think about it, you'd be better off leveraging edismax and >> the boost parameter. Don't mess with your main q(uery), use >> boost=genre:Citation^0.01 or something like that. boost params (not bq!) >> are multiplied into the score, not added. Maybe that'll be more to your >> liking? >> >> Erik >> >> >> On Oct 31, 2011, at 10:19 , Paul wrote: >> >>> Thanks Erik. They don't need to absolutely always be the bottom-most >>> -- just not near the top. But that sounds like an easy way to do it, >>> especially since it is a lot easier to reindex now than it used to be. >>> >>> I would like to know why my query had no effect, though. There's >>> obviously something I don't get about queries. >>> >>> On Mon, Oct 31, 2011 at 10:08 AM, Erik Hatcher <erik.hatc...@gmail.com> >>> wrote: >>>> Paul (*bows* to the NINES!) - >>>> >>>> If you literally want Citations always at the bottom regardless of other >>>> relevancy, then perhaps consider indexing boolean top_sort as true for >>>> everything Citations and false otherwise, then use &sort=top_sort >>>> asc,score desc (or do you need to desc top_sort? true then false or false >>>> then true?) >>>> >>>> Then you can have Citations literally at the bottom (and within that >>>> sorted in score order) and likewise with non-Citations at the top and >>>> sorted score order within that. Other tricks still risk having Citations >>>> mixed in should relevancy score be high enough. >>>> >>>> The morale of this story is: if you want to hard sort by something, then >>>> make a sort field that does it how you like rather than trying to get >>>> relevancy scoring to do it for you. >>>> >>>> Erik >>>> >>>> >>>> On Oct 28, 2011, at 17:17 , Paul wrote: >>>> >>>>> (I am using solr 3.4 and edismax.) >>>>> >>>>> In my index, I have a multivalued field named "genre". One of the >>>>> values this field can have is "Citation". I would like documents that >>>>> have a genre field of Citation to always be at the bottom of the >>>>> search results. >>>>> >>>>> I've been experimenting, but I can't seem to figure out the syntax of >>>>> the search I need. Here is the search that seems most logical to me >>>>> (newlines added here for readability): >>>>> >>>>> q=%2bcontent%3Anotes+genre%3ACitation^0.01 >>>>> &start=0 >>>>> &rows=3 >>>>> &fl=genre+title >>>>> &version=2.2 >>>>> &defType=edismax >>>>> >>>>> I get the same results whether I include "genre%3ACitation^0.01" or not. >>>>> >>>>> Just to see if my names were correct, I put a minus sign before >>>>> "genre" and it did, in fact, stop returning all the documents >>>>> containing Citation. >>>>> >>>>> What am I doing wrong? >>>>> >>>>> Here are the results from the above query: >>>>> >>>>> <response> >>>>> <lst name="responseHeader"> >>>>> <int name="status">0</int> >>>>> <int name="QTime">1</int> >>>>> <lst name="params"> >>>>> <str name="fl">genre title </str> >>>>> <str name="start">0</str> >>>>> <str name="q">+content:notes genre:Citation^0.01</str> >>>>> <str name="rows">3</str> >>>>> <str name="version">2.2</str> >>>>> <str name="defType">edismax</str> >>>>> </lst> >>>>> </lst> >>>>> <result name="response" numFound="1276" start="0"> >>>>> <doc> >>>>> <arr name="genre"><str>Citation</str><str>Fiction</str></arr> >>>>> <str name="title">Notes on novelists With some other notes</str> >>>>> </doc> >>>>> <doc> >>>>> <arr name="genre"><str>Citation</str></arr> >>>>> <str name="title">Novel notes</str> >>>>> </doc> >>>>> <doc> >>>>> <arr name="genre"><str>Citation</str></arr> >>>>> <str name="title">Knock about notes</str> >>>>> </doc> >>>>> </result> >>>>> </response> >>>> >>>> >> >> >
