The Future Enterprise Search Architecture
Continual growth in unstructured data volumes necessitates a change in the way enterprise search systems are designed:
- To ensure that documents can be indexed in a timely manner, without placing unacceptable load on IT infrastructure
- To maintain search results relevancy, despite the ever increasing volumes of content
Further, while addressing these key challenges, new enterprise search architectures, based on big data concepts, can also provide substantial improvements in search system reliability, cost-of-ownership, flexibility and operational agility.
In simple terms, it is necessary to view Ingestion Processing, and separately Content Processing prior to indexing, as largely autonomous sub-systems of the overall enterprise search architecture.
- Ingestion Processing becomes especially challenging as data volumes grow. Planning is needed to ensure that content can be extracted from enterprise repositories in a timely manner, but without overloading the repository and denying access to other systems and users. These are chellenges that plug-and-play connectors are seldom able to meet
- Content Processing must provide ever more sophisticated text analytics, normalization, cleansing and metadata enrichment in order to provide good raw material for relevancy algorithms to work with, and to drive user interface features such as search navigation, and alternative results sorting methods
For a full description of this new architecture, please read our free-to-download white paper, A Big Data Architecture for Enterprise Search
Search Technologies is working with a growing range of large companies to plan and implement future-proof enterprise search architectures. For further information, or an initial discussion of your circumstances, contact us.