In the first part of my blog on the Information Enrichment pillar, I discussed how autoclassification can be a power tool in providing your users a more meaningful search experience by tagging content with the right set of terms to drive findability and relevancy. In this follow up blog, I’ll discuss metadata expansion, which is another form of information enrichment.
Metadata expansion allows you to enhance the metadata associated with your content by adding additional supplementary metadata for the content from another content source or application. Let’s examine a common scenario. To illustrate, I’ll use a law firm as an example. The law firm has a document management system that has metadata such as title, author, date modified, client name, and the case name associated with the documents in the DMS. The law firm also has an application for client management, as well as another application for managing individual cases for the clients. The client management application has information such as the location, industry, and primary contact for the client. The case management application has information such as case status and court jurisdiction. The documents and associated metadata from the DMS is crawled and indexed using a connector from BA Insight which enables users to search for the documents based on the document’s content, author, client name or case name.
What if the users wanted to view or search for documents based on metadata in the client or case management application? For example, they wanted to search for documents based on a client location of Boston and a case (matter) status of open. This search use case would not be possible because the crawl of the documents in the DMS will only pull in metadata that’s available in the DMS and client location and case status is not available in the DMS. Using metadata expansion will allow you to set up such a use case. This is done by making calls to the supplementary metadata sources during the crawl of the documents from the DMS. In the above example, as each document is crawled from the DMS, a call is made to the client management application to get the client location based on the client name in the DMS. Another call is made to the case (matter) management application to get the case status based on the case name in the DMS. The client location and case status data retrieved from the respective systems are merged with the document’s metadata from the DMS and sent as a collective unit to crawl process as if all that information came from the DMS. The end result is that the document is indexed with the client location and case status in addition to the title, author, date modified, client name and case (matter) name. The users at the law firm are now able to find documents based on the client location or case (matter) status.
To illustrate the power of the metadata expansion process, below are two pictures that show what the search experience would be like without and with metadata expansion. The pictures below continue the scenario of the legal firm and the need to search for documents based on client and case (matter) related information.
This is what the non-metadata expanded search experience will look like. There is no way to search for documents based on client or case (matter) info because that information doesn’t exist in the DMS.
Now, here is the exact same search with the added benefit of client and case (matter) info being added to the available metadata of the documents. This not only allows the user to search for documents based on client name, client location, case (matter) or case (matter) status, but also allows refinement of the search results based on this enhanced document metadata.
I’ve oversimplified the above example to easily illustrate the metadata expansion concept. Metadata expansion is a pretty powerful mechanism to provide superior findability and relevancy that wouldn’t be possible with just the metadata from the original content source. This additional metadata can be used to drive refinement and also provide the visual cue your users will need to distinguish one search result from another. Metadata expansion is one of the key features of the BA Insight Connector Framework, which is the base set of features and functions for all BA Insight Connectors. The metadata expansion feature of the Framework is called Datasets, and more information about Datasets can be found in the online documentation for the Connector Framework.
To summarize information enrichment, crawling and indexing metadata from your content source is only the first step. Auto-classification and metadata expansion is often necessary to get the rich and powerful search experience your users will want for findability, relevancy, and refinement.
In the final blog of this series, I’ll cover the user experience aspect of building a killer search app. Stay tuned.