Haunted by Tchaikovsky: search woes

search
arabic
Tags: #<Tag:0x00007f23c72eacb0> #<Tag:0x00007f23c72ea850>

#1

So I’m adding an Egyptian release, which is hard enough, with me being illiterate in Arabic, and all.

Now I want to attach credits to a work, enter بيرم التونسي into the artist box, and who pops up in the top results? Pyotr Ilyich! You can see this easily here.

The above full search link shows six results with a score of zero. Why do they show up at all? This is especially problematic in the quick search boxes (e.g. in the relationship editor), where no scores are shown.

It seems like having a few letters in common is enough to get in the result list. There’s probably some of these in بيتر إليتش تشايكوفسكي (Tchaikovsky’s primary Arabic alias), though I couldn’t spot the similarities.


#2

It seems the parts بيتر (Peter) and بيرم (?) are quite similar and apparently causing this. Not sure what the Arabic name of this Mahmud Bayram el-Tunsi actually is; they don’t seem to match.

Perhaps someone who actually speaks Arabic can comment on this.


#3

بيرم (Bayram) is beh-yeh-reh-meem, while بيتر spells beh-yeh-teh-reh. Three matching letters (with one insertion) out of twelve.

Hmm, it is somewhat understandable now. Ok, sometimes the search just turns up dirt. As long as it’s scored correctly (which it was in this case), I’m not complaining (anymore).


#4

Those partial results are often quite good to have, as it also catches some typos and partial matches. You can also see this when searching for latin names. What makes it special in this case that we don’t have that any artist names in arabic script, thus resulting in an overall small list where the 0 score results look a bit too prominent maybe :smiley: