One of the most common issues we see at our clients is with the initial indexing of attributes for documents being imported into Documentum or Alfresco. This post will present the new “Heads up” indexing in HPI for both single and bulk import of documents.
Documentum or Alfresco – Indexing Approaches
Most interfaces try to emulate an “explorer” approach with the easy “upload” or “drag and drop” of files into a location within the repository. Once the file is dropped into the repository:
- The user might need to remember go back after the upload is completed and visit the attribute screen to add the metadata (Alfresco Share).
- The user might be prompted with the attribute screen to enter the metadata (Documentum Webtop/D2)
TSG’s HPI interface has always assumed that more than one document can be uploaded at a time, and before the documents are uploaded the user will be prompted to fill out the metadata for each document. You can see an example of simple bulk upload in our learning zone.
Many clients accustomed to scanning/indexing applications like Ephesoft have asked to be able to view the document while adding attributes. This is possible on an individual document by showing the property screen after the document has been uploaded.
This approach relies on the document being uploaded before setting metadata so that the PDF rendition of the document can be displayed. We find that many times users don’t remember to go back in and enter the metadata after the upload completes.
HPI Heads up Indexing with Bulk Load
Heads up indexing provides a “one step” approach for uploading/indexing and resolves the typical issues with uploading of documents by uploading the documents to a temporary location before prompting for the attributes. This approach works by:
- User selects one or multiple files via drag and drop
- Interface uploads files to temporary location and renditions them to PDF.
- Prompt the user for each object type and index components while viewing the document side-by-side with the attributes to be indexed for the selected document type
- Stores the document with attributes in the right location and deletes the temporary copy.
While the approach above is slightly slower as the documents need to be uploaded during indexing, for most documents the delay is slight and barely noticeable.
Heads up Copy and Paste
One of the big advantages of the “Heads Up” approach is the ability to leverage the text behind the PDF during the indexing step. By using our OpenAnnotate in an indexing mode, indexers can select words, sentences or whole paragraphs and see them quickly appear in the attribute fields.
This approach works well for any document. Some specific examples:
- Invoices – where the address, PO or other indexing values appear on the document
- Contracts – where whole paragraphs could be attributed to keep track of exposure of specific differences in contracts that should be tracked outside of full-text search.
A “heads up” approach for attribute indexing requires that the document be uploaded before indexing occurs. Most “out of the box” interfaces don’t handle this well as the document type and location need to be set before the document is uploaded. HPI resolves this issue by copying the files to a temporary space and then storing them in the final location after index values have been added while viewing the document. This approach also allows for index values to be “copied and pasted” from the document itself during the indexing process with a simple highlight approach.
Let us know your thoughts below.