Should pdf format e-books be searchable/have links?

Hello everyone,

My company has recently published a POD book called Paris Insights - An Anthology using Blurb.com. I have created a pdf file of this book by scanning the original. I have enabled the appropriate security setting for the file to discourage unauthorized editing and printing. I plan to use this file as an e-book, both for promotion and for sale on the Internet.

I converted a duplicate of this file to be searchable, but many of the words in the book did not convert consistently. Readers will be able to find a given word most, but not all, of the time when they use the search function. They may not be able to find a rarely used word at all if the scan of the text did not convert properly.

Also, the URLs that appear in the e-book do not have hyperlinks.

How important are these issues to the successful marketing of an e-book in pdf format? Should I plan to include a disclaimer with the file, saying that there may be problems with searching and that no direct Internet access to the web addresses in the text is available? Should I use the pdf file that is not searchable and mention this in the disclaimer? Or should I start from the beginning and create an e-book that has these features (something that I would prefer not to do)? Will anybody care one way or the other?

Thanks,
dpbooks

Thanks to both soma and veinglory for responding.

Have either of you used or heard of anyone using Amazon's Kindle software to create a searchable/linkable e-book from a pdf file? Might this solve my problem? Would any other program out there work for me?

dpbooks

Yes, that's correct.

The problem with the original file is that it has a Blurb watermark that says "for proofreading only" or something to that effect. I have used this function to print out copies for review. But the only way to get a "clean" pdf copy of the book for public distribution is to scan it.

Regarding the security features to prevent printing, I can easily remove this. But this will not help with 100% searchability or the ability to add links, correct?

dpbooks

I didn't know that this was possible. I will investigate and see what I can come up with.

Thanks very much for your advice!

dpbooks

Dear Elodie,

Thanks for informing me of this! I have edited my signature and the link works for me in Mozilla now. Would you please try it again to see if you can now read it with that browser?

dpbooks

I've got someone looking at the watermark question for me. But I now know that this is a moot point because I have found that the pdf file from Blurb is not searchable. OCR needs to be applied to this file.

Does anyone know if the application of OCR at the time of the physical scanning of a book works better than applying OCR to a file that has already been created? I already have a pdf file that is pretty well searchable, but that misses some words and shows others as gibberish. (BTW, I have successfully added hyperlinks to this file - there are not very many of them.)

In other words, would it be worth the time and expense to have the hard copy of the book rescanned, applying OCR at the time the scan is made?

Thanks,
dpbooks

Great!

Hi Soma,

I'm not sure that I understand what this would accomplish. The words are correctly spelled in the document (spell check was completed prior to publication). But when they are searched, words adjacent to the searched term can sometimes contain one or more symbols or an incoherent string of letters. Or, the searched term itself may contain an erroneous letter (for example, a search for the word "boat" will call up "about", with the letters "bout" being highlighted as the searched term. I think that this is because the scanner cannot distinguish between the letter "a" and the letter "u" 100% of the time.

If this is a limitation of OCR technology, perhaps there is nothing more that I can do. If you were in my place, would you use this file (hyperlinks included) as an e-publication? Would you use it for promotion of the printed book? I would say that the search function is working at about 95% capacity.

dpbooks

I see.

I am not familiar with this process. There are thousands of words that could be searched, and each search would generate a different set of words and phrases in the search box. How could I ever be assured that I have corrected all the words that appear abnormally?

Can anyone explain to me (a relative novice with manipulating pdf files) how to go about exporting the text found in the box when a word is searched and how to correct it? Or tell me where I can go to find the information?

Thanks,
dpbooks

Hi Soma,

I did what you suggested and exported the OCR document to Word. The result was a document that contains numerous different fonts and symbols instead of letters on each page, with all text rotated 90° on the page. Only the pages with photos and no text have retained the page color (light blue) and the correct orientation. The number of misspellings is enormous. It would be faster for me to start over again.

Last night, I read a chapter on e-publishing in a book by Penny Sansevieri called Red Hot Internet Publicity (Morgan James, 2007). Penny advises that an e-book should not be longer than 100 pages long (mine - or rather, my husband's - has 156 pages) and that it should have an "I need it now" element (my husband's book is a collection of essays on culture and contemporary life in Paris - compelling, we like to think, but hardly anything that shouts "urgent"). She suggests breaking the book into chunks and offer them separately to "drive more urgency than the book as a whole". What do you think of this advice?

Penny also says not to overload the book with pictures or graphics. But my husband's book contains 109 color photos - this is a major selling point of the work.

Perhaps we should abandon the concept of creating an exact copy of the printed book as an e-book and do something else altogether.

I welcome your comments and those of others reading this thread.

Thanks, and have a good weekend!
dpbooks

Do you suggest that we persist in trying to propose an e-version of the entire book then? Even if it is not a "clone" of the print version?

Okay, thanks anyway!

dpbooks