Re: Dev chat reminder, issue f6335fa6-517f-4cba-9083-7241e9d5d26a
Jeff Weeks <[email protected]> Mon, 24 Aug 2015 12:31:23 -0400
| Newsgroups | gmane.comp.audio.musicbrainz.devel |
|---|---|
| Message-ID | <CA+V91tgeyXQ+-BBuGE4kj6oDsARmFPYdFeEiWZx-kQ9UiKK6Tg@mail.gmail.com> |
--===============1155096870== Content-Type: multipart/alternative; boundary=001a113ecb4af82b1d051e112558 --001a113ecb4af82b1d051e112558 Content-Type: text/plain; charset=UTF-8 I'm not sure if I'll be around for the meeting today or not. I'm at the mercy of my auto repair shop today. I've been experimenting with edismax search configurations on the artist entity. https://wiki.apache.org/solr/ExtendedDisMax#Query_Syntax Right now I'm getting solid results. I was hoping to get results like "nirwana" when I searched for "nirvana" but my experimentation with ngrams hasn't really yielded anything useful. Between my own research (away from my development machine) and a chat on #solr this morning it has come to my attention that I could incorporate a fuzzy search on specific fields. Currently, my edismax setup looks for matches in the following fields (all equal weights and nothing fuzzy): alias, area, artist, comment, country, ipi, isni, mbid, ngram, sortname, tag, type The way I have ngram setup at the moment is to only return results where there is an exact ngram result to the query. So this returns "Nirvanananda" and would return "supercalifragilisticnirvanalidocious" but not "Nirwana". To get "Nirwana" I would need to analyze ngrams of the query (and this gives me a gazillion irrelevant results! like "Ludwig van Beethoven" ....hey, it contains "van"!). When I get home this evening, I'm going to try allowing fuzzy results on the alias, artist, and sortname fields that will allow results within 2 differences of the queried term. Other than the fuzzy difference, I'd love to get some feedback on my search results: (solr) https://www.dropbox.com/s/5hpfxmb7ercx3gy/solr-nirvana.png?dl=0 (current-mb) https://www.dropbox.com/s/s3jwrg9892mkv4h/mb-nirvana.png?dl=0 ruaok: this isn't up yet on my VM...just experimenting locally -weeksio On Sun, Aug 23, 2015 at 3:02 PM, Statler & Waldorf <[email protected]> wrote: > We've got our weekly dev chat on 2015-08-24 on IRC in #musicbrainz-devel > on irc.freenode.net. We're going to meet at Regular Meeting Time [1] > (19:00 UTC) [2]. > > If there is any topic you would like to discuss during the meeting, please > add it to the agenda in the channel topic. > > [1] http://musicbrainz.org/doc/Development_Chat > [2] http://www.timeanddate.com/worldclock/fixedtime.html?iso=20150824T19 > > -- > This message brought to you by https://github.com/mayhem/statler-waldorf > Don't even think of responding to this email. We won't answer! > http://goo.gl/FSZdF > > _______________________________________________ > MusicBrainz-devel mailing list > [email protected] > http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-devel > --001a113ecb4af82b1d051e112558 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable <div dir=3D"ltr"><div class=3D"gmail_default" style=3D"font-family:verdana,= sans-serif;font-size:small">I'm not sure if I'll be around for the = meeting today or not.=C2=A0 I'm at the mercy of my auto repair shop tod= ay.</div><div class=3D"gmail_default" style=3D"font-family:verdana,sans-ser= if;font-size:small"><br></div><div class=3D"gmail_default" style=3D""><font= face=3D"verdana, sans-serif">I've been experimenting with edismax sear= ch configurations on the artist entity.=C2=A0</font><br><font face=3D"verda= na, sans-serif"><a href=3D"https://wiki.apache.org/solr/ExtendedDisMax#Quer= y_Syntax">https://wiki.apache.org/solr/ExtendedDisMax#Query_Syntax</a></fon= t><br></div><div class=3D"gmail_default" style=3D""><font face=3D"verdana, = sans-serif"><br></font></div><div class=3D"gmail_default" style=3D""><font = face=3D"verdana, sans-serif">Right now I'm getting solid results. I was= hoping to get results like "nirwana" when I searched for "n= irvana" but my experimentation with ngrams hasn't really yielded a= nything useful.=C2=A0 Between my own research (away from my development mac= hine) and a chat on #solr this morning it has come to my attention that I c= ould incorporate a fuzzy search on specific fields. =C2=A0</font></div><div= class=3D"gmail_default" style=3D""><font face=3D"verdana, sans-serif"><br>= </font></div><div class=3D"gmail_default" style=3D""><font face=3D"verdana,= sans-serif">Currently, my edismax setup looks for matches in the following= fields (all equal weights and nothing fuzzy):</font></div><div class=3D"gm= ail_default" style=3D""><font face=3D"verdana, sans-serif">alias, area, art= ist, comment, country, ipi, isni, mbid, ngram, sortname, tag, type</font></= div><div class=3D"gmail_default" style=3D""><font face=3D"verdana, sans-ser= if"><br></font></div><div class=3D"gmail_default" style=3D""><font face=3D"= verdana, sans-serif">The way I have ngram setup at the moment is to only re= turn results where there is an exact ngram result to the query.=C2=A0 So th= is returns "Nirvanananda" and would return "supercalifragili= sticnirvanalidocious"=C2=A0</font><span style=3D"font-family:verdana,s= ans-serif">but not "Nirwana".=C2=A0 To get "Nirwana" I = would need to analyze ngrams of the query (and this gives me a gazillion=C2= =A0irrelevant results! like "Ludwig van Beethoven" =C2=A0....hey,= it contains "van"!).=C2=A0</span></div><div class=3D"gmail_defau= lt" style=3D""><font face=3D"verdana, sans-serif"><br></font></div><div cla= ss=3D"gmail_default" style=3D""><font face=3D"verdana, sans-serif">When I g= et home this evening, I'm going to try allowing fuzzy results on the al= ias, artist, and sortname fields that will allow results within 2 differenc= es of the queried term.</font></div><div class=3D"gmail_default" style=3D""= ><font face=3D"verdana, sans-serif"><br></font></div><div class=3D"gmail_de= fault" style=3D""><font face=3D"verdana, sans-serif">Other than the fuzzy d= ifference, I'd love to get some feedback on my search results:</font></= div><div class=3D"gmail_default" style=3D""><font face=3D"verdana, sans-ser= if">(solr)=C2=A0<a href=3D"https://www.dropbox.com/s/5hpfxmb7ercx3gy/solr-n= irvana.png?dl=3D0">https://www.dropbox.com/s/5hpfxmb7ercx3gy/solr-nirvana.p= ng?dl=3D0</a></font></div><div class=3D"gmail_default" style=3D""><font fac= e=3D"verdana, sans-serif">(current-mb)=C2=A0<a href=3D"https://www.dropbox.= com/s/s3jwrg9892mkv4h/mb-nirvana.png?dl=3D0">https://www.dropbox.com/s/s3jw= rg9892mkv4h/mb-nirvana.png?dl=3D0</a></font></div><div class=3D"gmail_defau= lt" style=3D""><font face=3D"verdana, sans-serif"><br></font></div><div cla= ss=3D"gmail_default" style=3D""><font face=3D"verdana, sans-serif">ruaok: t= his isn't up yet on my VM...just experimenting locally</font></div><div= class=3D"gmail_default" style=3D""><font face=3D"verdana, sans-serif"><br>= </font></div><div class=3D"gmail_default" style=3D""><font face=3D"verdana,= sans-serif"><br></font></div><div class=3D"gmail_default" style=3D""><font= face=3D"verdana, sans-serif">-weeksio</font></div><div class=3D"gmail_defa= ult" style=3D""><font face=3D"verdana, sans-serif"><br></font></div></div><= div class=3D"gmail_extra"><br><div class=3D"gmail_quote">On Sun, Aug 23, 20= 15 at 3:02 PM, Statler & Waldorf <span dir=3D"ltr"><<a href=3D"mailt= o:[email protected]" target=3D"_blank">[email protected]</a>>= ;</span> wrote:<br><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 = .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class=3D"HOEnZb"><di= v class=3D"h5">We've got our weekly dev chat on 2015-08-24 on IRC in #m= usicbrainz-devel on <a href=3D"http://irc.freenode.net" rel=3D"noreferrer" = target=3D"_blank">irc.freenode.net</a>. We're going to meet at Regular = Meeting Time [1] (19:00 UTC) [2].<br> <br> If there is any topic you would like to discuss during the meeting, please = add it to the agenda in the channel topic.<br> <br> [1] <a href=3D"http://musicbrainz.org/doc/Development_Chat" rel=3D"noreferr= er" target=3D"_blank">http://musicbrainz.org/doc/Development_Chat</a><br> [2] <a href=3D"http://www.timeanddate.com/worldclock/fixedtime.html?iso=3D2= 0150824T19" rel=3D"noreferrer" target=3D"_blank">http://www.timeanddate.com= /worldclock/fixedtime.html?iso=3D20150824T19</a><br> <br> --<br> This message brought to you by <a href=3D"https://github.com/mayhem/statler= -waldorf" rel=3D"noreferrer" target=3D"_blank">https://github.com/mayhem/st= atler-waldorf</a><br> Don't even think of responding to this email. We won't answer! <a h= ref=3D"http://goo.gl/FSZdF" rel=3D"noreferrer" target=3D"_blank">http://goo= .gl/FSZdF</a><br> <br> _______________________________________________<br> MusicBrainz-devel mailing list<br> <a href=3D"mailto:[email protected]">MusicBrainz-deve= [email protected]</a><br> <a href=3D"http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-devel"= rel=3D"noreferrer" target=3D"_blank">http://lists.musicbrainz.org/mailman/= listinfo/musicbrainz-devel</a><br> </div></div></blockquote></div><br></div> --001a113ecb4af82b1d051e112558-- --===============1155096870== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ MusicBrainz-devel mailing list [email protected] http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-devel --===============1155096870==--