Learning patterns/Making the Digital Library of India really useful
What problem does this solve?
The Digital Library of India is a repository of books, magazines, and other materials, mainly in Indic languages. But this became a locked room where the key was lost. This is because of the not so useful cataloging methods used by them. Although there are very useful books to Indic wikipedias, wikipedians can hardly use it because the books are having very wrong details of the title and author of the books.
What is the solution?
Create catalogs in wikipedias/wikisource in the respective Indic language linking every book entry to respective DLI page. So that the book titles and author names can be searchable. Involve the wiki community as well as the book lovers outside wiki.
- Create catalogs as tables with necessary categories: Creating catalog tables with necessary sections you've selected will be initial step to the great journey. Identify the useful and necessary categories, In this case I could say don't ever leave DLI Code. Because it gives unique identity to the given book. When you start working even if you miss giving link this could be greatly helpful to the end users.
- Choose your approach: As per our telugu wikimedian Arjuna rao chavala there are two different approaches to deal with this problem. First one is to make a bulk trail. If you choose this then rely on metadata and try to guess the right name of book according to the metadata and transliterate it to the original language from English. This will help you to work on more books within a short span. But the disadvantage is there are many instances to say these metadata may be completely wrong and maybe even no match with the real metadata inside the book. Another approach is to open the book and understand about it and create metadata in catalog. This may be a bit slow but results in an accurate catalog.
- You must open the book anyway: If you choose to make slow but mostly perfect catalog then never rely on the metadata given by the DLI website. Because there are many instances in which the book is entirely different and metadata is something else. Even Simhali, Urdu, Tamil and other languages books are categorized as telugu books. So, take pain and try opening the book, take a look at few pages and then work on cataloging that book.
- Work on-wiki: Working on-wiki is best approach in this project. We've tried to work off-wiki and create excel files and publish them online in many sites along with wikipedia. But after knowing it practically I got to know that working on-wiki has many advantages like not doubling the work, good co-operation between users, etc., So I suggest people to work on-wiki.