CIP D1 – Creating, Capturing and Sharing Information
Common Business File Formats
- PDF; TIFF; JPEG; MS Office (Word, Excel & PowerPoint)
Identifying Process Entry Points for Information
Process entry points include:
- Email/fax - Paper documents
- Voice/voicemail - Websites and web-based forms
- Internal and external workflows - Smartphones, tablets, & mobile apps
- Uploads to a file-sharing solution - Other channels to create or transmit information
Point of Service
One of the key considerations here and for multichannel capture more broadly is that, to paraphrase
AIIM research, we want to “digitize everything that moves” and do so as early as possible. Every step
that involves paper is less efficient than it should be. So ideally this capture and digitization happens
at the point of service or the point of the transaction. This is not a production imaging process, but
rather a very decentralized one.
Key Considerations
Regardless of the specific process entry point, but especially as applied to point of service capture
processes, there are some things to keep in mind to ensure the information received can be
processed effectively: completeness; format and quality; and chain of custody.
Determining the Best Points of Capture for Different Kinds of Information
Capture is the process of getting information from its source into some type of more formal information
management environment, system, application, or business process, and then recording its existence in
the system. You need to include relevant data about the paper document, other physical object or digital
file that has been captured. This ‘data about data’ is known as metadata.
Sources of Content
For the CIP, the focus is on enterprise information. Enterprise information can take any number of
forms. We may think of typical office documents such as spreadsheets or contracts, but any content
with business value can be worth managing in an IM system.
The Point of Capturing Information
The entire point of capturing information is to establish some sort of control and context around it
and ingesting it for automated processing in transactional processes. The most effective way to do
this is through formally capturing and storing information in some sort of repository.
It allows the organization to establish context around that information. What process is it part of?
Who processed it and in what fashion? When is it captured? What other documents, records, or
processes does it relate to? Who owns and who uses it?
Capture at the Point of Service
This will help to reduce transaction times and the overall costs of the process, making it more
efficient for the organization as well as for any customers or partners involved. If it can’t be captured
directly at the point of service, it should still be digitized (if applicable) and captured as early in the
process as possible. This should be the general rule regardless of the type or source of information.
, Point of Capture – Business Documents
- At the point of service or transaction.
- Automatically as part of a workflow, such as a review and approval workflow.
- As part of the authoring processes.
Point of Capture – Scanned Images
In a production imaging environment where hundreds or thousands of pages are being scanned
every day, the imaging process should include a step for releasing the images – to another process,
to a repository, or even to a file share location.
Point of Capture – Email
Email has been a challenging information type for decades because of very high volume, lot of junk,
informality, lots of topics, etc.
Point of Capture – Structured Data
In most organizations, structured data is “captured” into its own application. Every line of business
application has its own data tables, and perhaps its own database, and information is captured as it
is entered into the system. In some cases, this structured data is used to generate reports, which
may in turn need to be captured and managed. These may need to be captured manually or through
a workflow of some kind to ensure all the data is present. Structured data can also be captured more
transactionally through the use of forms. In the case of paper-based forms, the most efficient way to
capture this data accurately is through scanning the forms and then using recognition technologies
to automatically extract and capture that data. Digital forms may have their data captured directly
into a structured line of business application.
Introduction to Document Management
It is the use of a software application to track digital documents from creation through approval and
publication. Traditional document management includes the following capabilities:
- Check-in/check-out : While the document is checked out, nobody else can edit it, and,
depending on the solution, it may not even be accessible in a read-only mode.
- Version control : Version control is used to manage or control different versions of a document
as it goes through the authoring and approval process. New versions are automatically created
through auto-save, by saving the document manually, or, in this specific context, by checking the
document back in.
- Roll Back : Many systems that offer document control capabilities offer the ability to roll back or
revert to the previous version. This is often done when a version is released prematurely or with
some sort of error.
- Security and access controls : Help to ensure that any changes made to a document are done
only by authorized users.
- Audit trails : Audit trails show what has happened in a system. In the context of document
management, audit trails can track every change to a document throughout its lifecycle,
including who made what changes, when, and in what sequence.
Systems of Record
It as “an information management system that is the authoritative source for a given data element or
piece of information.” This is not a recordkeeping system per se, although it can be. Rather, this is the
place where a particular type of information is stored. Ultimately, the organization should have a system
of record identified for every type of business information it creates, receives, and manages – i.e., “a
place for everything and everything in its place.”
,The Capture Process
The Paper Capture Process
The digitization process includes several sub-processes:
The Born-Digital Capture Process
You will need to make a reasonably quick assessment of exactly how digital documents and records
are created, received, and used and where the key risks are in terms of information that needs to be
captured.
Once you have identified your potential risk areas, you will need to examine the organization and its
work processes to determine exactly what types of information need to be captured. By doing this,
you will also gain a much better idea of the business transactions that take place in the organization
and the processes for creating and using information.
Next, you need to determine when in the process to capture a particular type of information. It is best
to capture information at the earliest point of entry into an organization to allow for automated digital
processing. After all, capturing drafts and capturing final documents or records would occur at
different points in the creation and publishing process.
The Capture Format
The organization will need to determine in what format that information should be captured.
- Original native format: This is the preferred approach because it is the most faithful
representation of the transaction. In case of legal.
- Commonly used format: For example, latest version of Adobe Acrobat.
- Standards-based format: This approach requires users to save the record into a non-proprietary
format such as plain text, HTML, or PDF/A. This is better for ensuring long-term access.
Why Manual Capture Doesn’t Work
In a majority of organizations, the majority of users do not identify and capture and manage their
information properly. Reasons: to high workloads and no time for it; lack of knowledge and training;
incorrectly classifying documents; the possibility of an error.
Automating Capture
There are a number of ways that the capture of born-digital information can be automated.
- Bulk import: Often when migrating information from one location/system to another.
- Workflow: For example, once the final document has been approved, it can be converted to PDF,
the PDF posted to the website, and the original finalized document declared a record and moved
to the records repository.
- Content types: This approach provides for the definition of characteristics that define a certain
information type such as invoices, contracts, etc. Every document that meets that definition is
automatically processed in a certain way.
- Analytics. This approach leverages machine learning to understand the contents of documents
or records and process them according to that understanding.
, Approvals
The final step in the quality control process is approval that the images were captured correctly. This
could be done through a review and approval-type workflow, or through a manual process. For both,
it’s important that all parts of the document are complete and accurate, and that supporting
documentation that is required has been captured as well.
Auditing the Capture Process
It is important to audit the capture process periodically to make sure that information is being
captured accurately and consistently. This audit needs to include the documents or records
themselves, but it should also include metadata.
Capture Process Metrics
Here are some capture process metrics that can be captured as part of the audit process. These
metrics may help make the business case for effective information management.
- The percentage of information captured into a particular repository – documents, records, and
metadata.
- Growth rates in terms of volume and in the rate of growth. This will help the organization plan for
future storage needs.
- Access and retrieval rates – how much information is actually accessed and used.
- The percentage of information that has been captured but no longer has business value – or its
value is unknown.
Requirements for Multichannel Capture
Multichannel capture is capture of different types of information from a variety of sources.
Types of Multichannel Capture: Multichannel capture also describes the point and approach to capture.
It certainly includes traditional production capture but extends beyond that to include ad hoc capture at
the point of service or of a transaction. And it includes on demand capture, often using a multifunction
device or mobile device.
Classification and Routing
Classification and routing can be done through static rules and template-based recognition. The less
advanced approach is based on keywords and rules and regular expressions. The more advanced
approach includes technology like Machine Learning and AI, Natural Language Processing and
Semantic Analysis. Solutions need to be trained on specific document types, which may require
large training sets, which are difficult to collect. The latest AI tools require significantly smaller
training sets.
A key piece of the capture process is assigning metadata to the document. This can be automated as
well using many of the same techniques such as data extraction, i.e., identifying the relevant data
fields and providing them to the respective business process.
Security
Security is even more important when digital information can come in from so many sources. As
things come in, the organization needs to implement effective processes and protocols to ensure for
example that personal medical information doesn’t sit on an open fax or in a shared inbox.
Incoming digital documents need to be scanned for viruses and malware, too, to make sure they
don’t cause information security issues.
And access controls need to be set up to ensure that only authorized users have access to the
incoming information, and only for the purposes they need it for.
