Metadata is usually seen mainly as a security risk: do you want people finding out who worked on a document and when, or what changes they made? But it's also a powerful way of tracking down information; after all, metadata such as the keywords on the page and the number of links to it are part of what Google uses to find you the web page you are searching for.
In the office, you can work out which files to archive by looking for all the documents more than two years old that no one has opened in the past six months. Better still, that is all free information: because applications record a lot of metadata for you, you don't need to train people to fill it in correctly, or rely on them making time to do it.
The problem is how few applications make the metadata available in a way you can work with. Microsoft Office captures plenty of information, and you can add custom properties to make it more sophisticated. But you have to look in the File Properties dialog or run a search for matching documents inside Office, or use extra software such as Exchange, SharePoint, a third-party product like JITI Builder or something you build with VBA or the Windows Indexing Service, to get at the data.
The XML support in Office 2003 will make document contents and metadata much more accessible if you build a system that uses the XML features. But if you want an example of how powerful metadata can be when software makes the most of it, take a look at the metadata features in the recently announced Adobe Creative Suite.
The new versions of Photoshop, Illustrator, InDesign and GoLive will all use Adobe's Extensible Metadata Platform (XMP) system to attach metadata to the file in the same format (on Mac and PC). Look at images in the Photoshop CS file browser and you can see as much or as little of the metadata as you want: place the same image in an InDesign document and get the same information in a tool tip when you hover your mouse over it.
Creative Suite includes a new tool that works in all the applications, called Version Cue. Designers frequently create multiple versions of a document to compare different effects: Version Cue lets you save those versions in the same file and label which is which - using XMP metadata. If your company logo changes, you can save the new one into the same file, labelling the up-to-date version so everyone knows what they should be using, but still leaving the older version available for old documents, if necessary. Labels, comments, keywords, copyright information: anything you fill in becomes metadata that is available in Version Cue, and in other ways, too.
Version Cue is mainly designed to improve workflow - whether you are trying to keep track of your own work or coordinate with a team. You see the information in the standard File dialog boxes. Even if you don't fill in any descriptions, the software still saves the metadata, such as the time the new version was created, so you can look for the version you were working on last Friday and compare it with the version you did on Monday morning.
That kind of metadata is most useful to the people working with the files, but metadata such as copyright information for an image is useful throughout a business. Adobe is working with system integrators and partners such as Documentum and Corbis to get XMP support into document management systems and image databases so the metadata can be used to generate reports or automate billing.
One newspaper Adobe has been working with uses metadata to make classified ads more profitable. The later you can include adverts, the more you have the chance to sell each issue, so the more automated the production and billing process, the better. Advertisers send their credit card details in with the advert; the billing details are saved as metadata, as is the number of characters - which tells the automated billing system how much to charge them without any manual processing.
If you are using a document as evidence, it is vital to know what has been done to it - and to be able to prove it. Adobe has also been working with UK police forces to document photographic evidence using metadata; the Nikon digital cameras used have GPS receivers so the metadata includes the place and time the picture was taken and even which memory card was used in the camera.
Previously, the police had to burn a copy of the original image to CD as a reference before doing any work to enhance or label it. Now they can open the Raw format image directly from the camera and track every manipulation they make to the image. Those changes (which you'd normally see in the history palette and lose when you close the file) are written to a text file and stored in the XMP packet. They can add custom metadata, such as the case ID, and transfer the image into a secure PDF template that prints the image as well as all the metadata. The US government processed all the footage from the Iraq war in Premier and is looking for ways of using metadata to add more levels of security and documentation.
As electronic evidence becomes more important in the courts, the integrity of the metadata in the evidence is going to be critical: an "enhanced" image could easily be misleading, but if you know it has been changed, you can demand the original. Logging the image and tracking the editing history is a step in the right direction, and could be useful for institutions worried their software might be used for working with inappropriate images.
For the rest of us, it means we are likely to see sophisticated tools for dealing with metadata that make it faster and easier to get the right file at the right time.
