Searching inside sealed documents
By Simon Thorpe on Jul 11, 2008
Organizations have many documents exposed on network shares, confined within content repositories and freely residing on local computers and storage devices such as USB flash drives. The growth of peer to peer networks, social websites and instant chat applications also make it far easier than ever before to share information quickly and easily outside of traditional enterprise network boundaries.
Searching across all of these environments for your information using Google-style full text search is now a crucial part of the digital working day. Technologies such as Oracle's Secure Enterprise Search enable enterprises to index and search applications, databases, file repositories, email archives and much more.
These same organizations are also increasingly using Information Rights Management (IRM) solutions to protect sensitive information. However the encryption used by these systems typically prevents full text searching of the information. So how do you ensure your sensitive documents and emails are secure and yet at the same time continue to allow authorized users to search for information they should have access to?
Being unable to full text search for information inside secured documents would cause the business serious pain. For example, in legal discovery, internal legal teams would be unable to search for documents requested during litigation or audit processes. Large fines are already being imposed on public companies that cannot produce information in specific, often very short, timescales. Another example would be an engineer searching for documentation containing valuable intellectual property. Being unable to find the right information would delay the progress of expensive engineers and be a hindrance to the efficiency of the company.
The Oracle IRM team has long recognized the importance of enabling existing search technologies to work with encrypted (“sealed”) information (subject to a users rights) and is the only IRM vendor to have developed Trusted Search™ products, now in use with many of its customers. The freely available Oracle IRM Desktop client, required to create and use sealed documents, includes support for full-text indexing and search within sealed documents using existing Windows search tools. This patented Trusted Search™ functionality is also available to be integrated with other Oracle products such as Secure Enterprise Search, and is available in the form of programming APIs for you to integrate with your own indexing and search systems.
The screen shot below shows a search result across a folder containing four sealed documents. Note how the search query only returned the engineering test report even though the other documents contain the string "Hitachi". The other files are protected against a classification to which the user does not have access.
In the latest release of Oracle IRM, 10gR3, we have also integrated this search functionality with the Oracle Enterprise Content Management system. All content that is checked in to the content repository and sealed is automatically indexed. Adding another out-of-the-box feature in the ability to search across your enterprise for valuable information even when it has been secured using IRM. Using the same files as the above example, the screenshot below shows the result of a search query inside the content repository. This same functionality has also been implemented by some of our customers in Documentum and Microsoft Sharepoint with the help of some basic consulting services.
In contrast, competing solutions from Adobe and Microsoft do not support full text search of their encrypted documents, and expect the customer to develop their own search engine integrations. Liquid Machines has integrated with Google's enterprise search appliance, the Google Mini, but in practice this is not widely used within enterprises, requires changes to IT infrastructure and end user workflows, and is tailored mainly to web based content.
If you'd like to learn about this first hand, contact us for an evaluation of Oracle IRM.