Translation Tips - OCR to Develop a Strategy Before Starting

Optical Character Recognition (OCR)

And if you do plan to use a translation memory tool, you will definitely also want a good OCR program. Which can prove quite useful even if you are not using a translation memory tool. For example, you get a large .pdf file which has a lot of tables and graphs in it. One option is to have that file open in one program and visible in the top half of your computer monitor, as explained above, while typing into a blank Word file in the bottom half of your computer monitor. With this approach, you will have to create the tables, possibly punch all the numbers in manually, and possibly resort to some fiddly approach concerning the graphics.

But what a good OCR program does is it scans the original document (you can use a scanner for paper documents, or simply open the file in the OCR program if it is an electronic document), creates the tables for you, punches in all the numbers, then converts the text into actual text in a Word document, or any other format that you request it to. So you can go right away and just type over the text as explained later and not have to worry or waste your time with creating tables and punching in numbers.

Or you can take this new electronic version of the original document and import it into your translation memory software, so that repeating phrases are automatically translated for you, and you do not have to worry about formatting at all, exporting your translation into the original format once you are completed. Isn't that wonderful?

There are many OCR programs but the best seems to be FineReader. It can recognise many languages too, and the latest versions are very good at the most complicated formatting, and can import from many different types of programs.

Develop a Strategy Before Starting, and Writing Over Top of an Electronic Document

If you will not be using translation memory software but rather type directly over top of an electronic document, there are certain tips which can help you with this.

First of all, it all depends on what you receive from your customer. After all, if you deal with direct customers, you can expect that, after you read all these fantastic tips, they will be hardly as savvy as you will be in the fine art of different translation strategies. So generally it is a good idea to take a look at a document and spend some time analysing it a bit before starting. The few seconds you save lunging into a translation could cost you dearly later, so take the time to decide on a good strategy. For example, after perusing the document a bit, you might find sections which will be extremely difficult for you.

Perhaps in a specialised field which you will not be able to handle. It will not go well for you if you translate 90% of a document, then realise that there are sections you cannot do, run out of time, and have to tell your customer of this problem around the deadline. You can imagine that your customer will be very angry, will have to find someone else to complete it, get it done late, and probably not want to pay you much more than half of the amount you did translate, if anything at all. Not to mention that they will probably never want to use your services again, because they cannot trust you. So before accepting any work, make sure you look through the document carefully and seriously consider if you can do a good job with it. If you tell your customer it is too hard for you, or you even help them find someone else, they will be happy that they can trust you, and it would be better than doing a bad job. If and when they find your mistakes, they will simply never trust you again.

Once you have accepted the work, you should take a little bit of time to examine it. You might be surprised to find entire sections which repeat or are very similar in content. In such as a case, there are several ways how you can deal with this, explained later.

If the document has a lot of tables and graphs, you might consider OCRing. If it has a lot of text which looks like you have translated in the past, you can pump it into your translation memory software and reap the juicy rewards.

It might be easiest to just prepare it from scratch in an empty Word file, but for this you will need to know how to format in Word, explained in the next section.

And the last option could be to type directly in the document, whether it was sent to you in electronic format, or you converted it into that with your OCR program.

One option is to get an upgraded version of Adobe Acrobat (to which you can find a link on our Download Translation Programs pages) which would allow you to type directly in the .pdf file.
A lot of times your customer might want the translation delivered in Word, so that they can use it for their internal needs and because their secretaries and other staff are not readily equipped with a fancy program like you have and which allows you to edit .pdf files. In this case, the OCR program can convert the format to Word.