Content Engine supports content-based retrieval (CBR) for documents, annotations, folders, and custom objects or their properties. With CBR, you can search an object store for objects that contain specific words or phrases embedded in document or annotation content or embedded in string properties of objects that have been configured for full-text indexing.
Each object store that is configured to support CBR requires the creation of full-text indexes. Content Engine creates and maintains these full-text indexes using the FileNet P8 Content Search Engine. Content Search Engine uses technology from Autonomy called K2 Enterprise. The full-text indexes are stored in index areas separate from an object store's database. An index area is represented by an IndexArea object in the database, and the IndexArea object contains the name of a file system directory to hold the full text indexing data.
For information on enabling full-text indexing, see Enable CBR for class definitions and properties.
You can customize the indexes to support searches that exclude "noise" words, are case-insensitive, or recognize sentence and paragraph boundaries by modifying the Verity Collection files. Once the indexes are created, they are updated as objects are added or checked-in to the object store.
Content-based searches can be performed from Enterprise Manager, Workplace XT, or other FileNet P8 clients. CBR searches performed from Enterprise Manager are limited to one object store at a time. However, CBR searches performed from Search Designer can search multiple object stores at a time. The result set is merged into a single normalized and ranked set.