Oibane, Indexing the Quran is exactly what I wanted to do. However, I cannot do this because, my arabic is very weak , I'm not a speaker! Anyway, I came across on a site somewhere, mentioning that in Quran, there are about 2000 distinct words only. So, I think it would not be that bad to index 2000 words. In fact, www.openburhan.com have done even further than that; it even have index based on root words!
I was thinking of integrating my work with openburhan (the database is with me), but I'm lacking both ideas and the time/energy to proceed. But note that openburhan does not use orignal Quranic spelling. If anyone have some ideas on how it should be done and willing to contribute, let get it rolling ... Regards. On 12/30/05, Oibane <[EMAIL PROTECTED]> wrote: > Seems like we're becoming close to the answer now. > > Generally speaking, putting aside Qur'aan, searching inside, say > a book written by someone, can be devided into two extreme cases: > google-like letter-by-letter one, and keyword/thesaurus based > index searching. > > If one is to use the literal matching, they should much or less > understand the author's phraseology or if in the language which allows > mutiple spellings, also spelling habits. There's more, difference of > meaning by locality, time, etc. > > As for a keyword index, first, let's imagine an index of a book. Since > we're in the cyber age, it can be integrable into literal search > engine, so that users can search, being unaware of the existence of > the index. You should pick up keywords first. Then relation between > terms are also exhibited in a good index. So under the heading of > "computer", there's sub-entries like "computer science", "computer > virus", and it reads "see also server, PC, calculator". Maybe behind > it is a tree-like diagram from general words indicating genre down to > more specific ones. > (I don't know the name of such study. It is related to bibliography.) > > I have never experinced it, but indexing appears to me a real > demanding task. Even if aided by PC power, it consists of human work > and work and work. But those who wish to explore through the > book will benefit greatly. And by indexing, you understand the book deeply. > > Now forget all what I said above, and come back to the Qur'an. Since > it is al-Qur'an, it has nothing to do with a particular novelist's > writing peculiarity. Yes, it IS the one. But not so simple; it is surely > difficult to search through for many people. > > One thing clear is that there're Arabic-speaking users, and > non-speaking, probably muslims. Needs vary among them. > > Perhaps clarifying the next coming job would reduce the complexity. > What Nicholas (Heer) and Mete referred to appears a high peak to climb > for the first time, although marvelous once accomplished. And if it is > the one to tackle with, it may well be worth an independent new > project, related but outside of "Qur'an" project, because it is quite > applicable for other classical materials. I think your concern on yeh > and hamza can establish a "better-then-google". > > Good afternoon. > > "Oibane" > > > _______________________________________________ > General mailing list > [email protected] > http://lists.arabeyes.org/mailman/listinfo/general >
_______________________________________________ General mailing list [email protected] http://lists.arabeyes.org/mailman/listinfo/general
