Full-cycle Methodology

Business Semantics Management Methodology

Adopt full-cycle Business Semantics Management to achieve Semantic Alignment

By default, business semantics serve “open” information systems, and hence the requirements and limitations of semantic alignment cannot be entirely known before completion. In contrast to waterfall-like approaches that focus on a broad design upfront, agile methods perform short milestone driven revision iterations in order to cope with dynamic environments such as the extended enterprise. Collibra’s full-cycle business semantics management is established by two operational cycles each grouping a number of activities.

Semantic Reconciliation

Semantic Reconciliation is the first cycle of the full-cycle business semantics management methodology. In this phase, business semantics are modeled by extracting, refining, articulating and consolidating fact-types from existing sources such as natural language descriptions, existing metadata, etc. Ultimately, this results in a number of consolidated language-neutral semantic patterns that are articulated with informal meaning descriptions (e.g., WordNet word senses). These patterns are reusable for constructing various semantic applications.

Semantic Application

Semantic Application is the second cycle of Collibra’s full-cycle business semantics man- agement. During this cycle, existing information sources and services are committed to a selection of semantic patterns. This is done by selecting the relevant patterns, constraining their interpretation and finally mapping (or committing) the selection on the existing data sources. In other words, a commitment creates a bidirectional link between the existing data sources and services and the business semantics that describe the information assets of an organization. The existing data itself is not moved. On the contrary, the business se- mantics provide a kind of abstraction layer to access and deliver this data in a more efficient and aligned manner.