Quick Fields 8 includes several processes designed to help you process electronic documents by working with their associated text.

Example: The Wonderland City Council uses Quick Fields to process their meeting agendas, which are created as and stored in Laserfiche as PDF documents. Quick Fields extracts the meeting date and some other information from the agendas, places this information in fields, and uses it to name and file the agendas in Laserfiche.

First, make sure the documents have associated text. With PDFs, you can confirm this by checking that you are able to highlight and copy the text on a document. You will be able to view PDFs within Quick Fields if you have Adobe Reader installed on the same computer as Quick Fields. Other types of documents can also be processed by Quick Fields, though they cannot be viewed within it.

Text Extraction
The first step is to extract text from the electronic documents.

To generate text from electronic documents in Quick Fields:

1. Select the Pre-Classification Processing node and select the Text Extraction process (it is only available in this stage of processing).
2. Select the types of documents to extract text from. If you select the Additional File Types checkbox, you can see a full list of available electronic document types.
3. Specify the number of lines that will constitute a page of the extracted text. Note that the lines of extracted text may not correspond to the appearance of the text on the electronic document.

Text Identification
Once you have extracted text from the documents brought into the session, you can create a pattern that will determine whether the documents belong to a particular document class. You can use a Text Identification process, or a Pattern Matching process, which can search selected lines only. The best practice is to identify according to something on the first page of the extracted text, to avoid splitting the text pages into separate documents.

To configure a Text Identification process:

1. Select the Identification node within a document class and select the Text Identification process (it is only available in this stage of processing).
2. Select Add Pattern and specify a pattern to match.

Example: The Wonderland City Council is identifying documents that contain the phrase Regular City Council Meeting as belonging to the document class named Regular Meetings.

Pattern Matching
You can use Pattern Matching to search the document text for information that meets specified patterns and generate tokens that can be used to fill out fields, folder paths, and document names.

To configure a Pattern Matching process to extract data from document text:

1. Select the Page Processing node within a document class and select the Pattern Matching process (it is also available in other stages of processing).
2. Under Page Range, specify the pages to be processed. Note the relationship between this setting and the Lines Per Page setting in Text Extraction.
3. Under Pattern Configuration, select Create Token. In the New Token dialog box, select the option to look for a pattern in page text. Specify the entire page or specific lines to search.
4. Specify a pattern or patterns to search for.

Example: The Wonderland City Council is searching for two patterns. They are searching lines 1-4 for a pattern that matches the format of a date. They are also searching the full text for the phrase “Ordinance No.” and capturing the number that follows it.

5. You can place the tokens generated by this process in fields, document name, or folder path so they will be filled out with the information generated when the documents are processed.

For more information on pattern matching syntax, see the Pattern Matching and Token Formatting in Quick Fields 8 white paper.

• Link this:
• Print this article

Posted on Monday, December 28th, 2009 at 12:00 pm under Support. You can comment on this post.