Question

Import batch of scanned back files

Scanning Capture

Updated February 3, 2016

asked on January 28, 2016 • Show version history

Hi,

I'm working with a client who has a new Laserfiche system and needs to import a large batch of newly scanned back files.

How does indexing these back files work? Is the scanning bureau supposed to produce some kind of metadata file (ex. '.lst' file) or some other output that can be used by Laserfiche to apply to template fields during the import?

Additionally, who initially provides the metadata? Is the scanning bureau supposed to produce it? Or is the client supposed to have some kind of indexing info or metadata before sending their files to the scanning bureau?

If anyone has experience with this, please let me know, it would be greatly appreciated.

0 0

Answer

SELECTED ANSWER

replied on January 28, 2016

What metadata currently exists on the boxes, or if you are lucky, on folders in the boxes?

When we had a service bureau perform a similar backscan project for us on approximately 100,000 docs in 700 boxes, we had them collect the data from the box/folder naming convention that had been used to file the paper originally, and then create a CSV file for us that would match the filename of the tiff scan for each doc to the box/folder info. It wasn't much additional metadata, but it helped.

I would suggest giving some thought as to how users will search for these docs once they are in Laserfiche. If you have limited or no metadata you really need to OCR everything as you import. (Use DCC if you have lot of docs.) At least then you'll have a fighting chance to find a document later. For us, OCR made a dumping ground for old docs something actually usable.

2 0

Replies

replied on January 28, 2016

Hi Nareg

It's all based on whether you customer requires MetaData to be assigned to the documents. If they have a Windows directory that you want to upload, I find that Import Agent works well, as it can be configured to maintain the current windows folder structure and file naming.

If there is Metadata that you want to upload at the same time, as the files may have come out of a legacy system, then I have used a CSV file (FilePath, FileName, Template, MetaData) and workflow to upload the documents and assign the templates and MetaData at the time of upload. If you have a very large quantity of documents this may not be the most efficient, at which time you may look at using SDK or a Migration Service.

1 0

replied on January 28, 2016

To Geoff's point, there is where you need to think this out before starting to scan. If you do feel that MetaData would be important\helpful then you need to make those decisions upfront. Either way it requires somebody to sort through the documents and follow the system.

Without knowing what Scanner application is on the front of you system it's hard to get too detailed.

One option is to use a naming convention or file structure such as Customer number_name_Descriptor, which the scanner operator would follow to name\file each of the documents. You could later use this information to draw\assign the Metadata

Secondly would be to implement Separator\Scanner Sheets that may have metadata on them. You create a template in Word that you change for each customer to put at the front of the stack.

Once the documents are created and imported into Laserfiche, you could use Workflow at this point to create the Metadata

1 0

replied on January 28, 2016

Hi Steve,

thanks for the info. The client's files are currently in boxes, so no legacy system. We're going to help them with scanning services so they get their paper files into digital format, then we'll import those into LF.

My question really is if it's even possible to produce metadata in this scenario, where the back files are coming from newly scanned papers documents, and not a legacy system.

0 0

replied on February 3, 2016

Thanks to everyone for replying, your answers definitely helped.

I asked the client a bunch of questions regarding the condition of their files, naming/filing convention, b/w vs color, number of files, etc. They provided limited information. I am working with a scanning bureau at the moment and it seems they have a good plan as to extract metadata during the scanning process. The client isn't even sure yet if they want metadata or not. But I appreciate everyone chiming in. Thanks!

0 0

You are not allowed to follow up in this post.

Question

Question

Import batch of scanned back files

Answer

Replies

Sign in to reply to this post.