Hi Everyone, this question may seem a bit below par but I am really struggling. I have created 10 or more Document Definitions each for individual customers with individual appropriate fields. Issue is I want to run the Capture in a typical machine format with little human involvement. Question is how to set it such that best possible match average .65 and above match the DD and extracts the data for export without hindrance. My approach thus far has been creating a new DD for each customer and then selecting option to use best possible DD. This has not worked successfully. I managed to export stuff to a datafile but the data was inconsistent. Typically the Capture machine will pull files from a common hotspot, capture the data and then export it to a database, however on pulling it needs to recognise that it is dealing with a defined document and match accordingly. Is this possible anyone??
How to export data from more than ten Document Definitions
この記事は役に立ちましたか？
0人中0人がこの記事が役に立ったと言っています
コメント
9件のコメント
I don't quite understand what you're trying to accomplish. Are you a BPO where you process other people's document for them? So you want to separate out the DocDef for each customer? If so, why not just use a different Batch Type for each customer. This way you can select only 1 DocDef to use for that Batch Type and it should match any other DocDef. Of course in this case, you will need to have different hotfolder for each customer.
If you still want to use one hot folder and use the default Batch Type (It will use all the DocDef in the project), then you need to improve the DocDef. Why is are they not matching correctly? Do they look similar except for one section? If so, just use that section as a required static text.
HI Sushi we are an organisation intent on automating our order handling process. We have many clients above 300 and they each have a different image/pdf/document. I have created the Layout files for them as well as the Capture files now I need to get them in the one hotfolder where they will be picked and selected according to the definitions I am guessing this is all done by the classifier. I tried the second recommended solution above but it took the definition of the first and applied to all the other images. The images all look very different however the same image from a particular customer can appear 4 or six times randomly in the hotfolder as they are being read into the hotfolder via email. I think the hotfolder solution will be successful but is that the most efficient solution with ABBYY?
Harti,
This is going to be a tough one if you can't pre-separate the images to their own batch type. The reason is because, with 300 DocDef to deal with, you will have issue with performance issue. It might take a couple mins to just ID 1 page. If they MUST all come in from 1 hotfolder, what you can do is use a description file to preselect the batchtype it should go into. If you can generate an XML and tell it what file to pickup and which batch type to select it should still let you use Batch Type. You don't technically need to create a classifer project if you want to use one hotfolder but it could help with the performance if you do.
Basically I'm trying to prevent you from having performance issue. I've dealt with customer that had over 50-100 FlexiLayout, and even with a faster computer, it was very slow. We ended up using batch type to resolve the performance issue.
Hi Shushi many thanks your suggested method of solution was very helpful. The main problems we are now experiencing is how to automate the verification when importing from hotfolder. By enabling auto verification we aim to truly get automated.
What do you mean by automate verification? Generally, after Recognition, the documents go automatically to Verification if there are any issue.
Hey Sushi, it turns out that in 98% of all the imported files there is always a verification pending after recognition. We need it to auto do this in order to export after?
This is basically setup in the FlexiCapture Workflow of the project. Depending on the issue, it actually might be best to verify the information. If the information is actually accurate, then you can lower the threshold for it in the Verification tab of the field.
Btw the way, have you ever gone to training for the product. A lot of your question are really basic question. Check with your local office to see if they have training. My local ABBYY office offer a monthly online training that helped me a lot.
Harti,
I don't understand what you mean by the Automation of the Recognition? It should happen automatically. Are you doing some sort of custom workflow?
Also this is why I recommend going to the training if there is one provided in your area. The documentation that comes with this product is a bit overwhelming if you don't get familiar with the product. It makes more sense now once I attended the online training. I wouldn't say the information is lacking, more of too much information and not organized for the basic user.
サインインしてコメントを残してください。