Knowledge Modeling includes the explicitation of Knowledge and Requirements that is available in Documents, such as design manuals, (international) Standard Specifications and Standard Data Sheets. In order to make such knowledge computer interpretable, it needs to be expressed in a formal knowledge representation language and thus transformed into a computer interpretable form. Knowledge Modeling also involves inputs of people and content, but more emphasis is on stakeholder/user input. Content contains information, but people contain knowledge, so knowledge modeling requires the input of various people, with the input gathered in a comprehensive and systematic way, such as through interactive brainstorming workshops and interviews. Knowledge Model is more similar to a knowledge organization system.
Semantic Modeling is a type of knowledge modeling that describes a network of concepts (ideas or topics of concern) and their relationships. Incorporated into information systems, semantic models enable users to ask questions of the information in a non-technical way. For example, a semantic model can map database tables and views to concepts that are meaningful to business users. Semantic models contain semantic objects and bindings. Semantic objects are things represented in the model. They can have attributes with cardinality and domains, and identifiers. Their structures can be simple, composite, compound, hybrid, association, parent / subtype, or archetype / version. Bindings represent associations or association classes in UML. These models help to identify patterns and trends and to discover relationships between pieces of information that might otherwise appear disparate. In doing so, they help enable integration of data across different knowledge domains or subject areas. Ontologies and controlled vocabularies are critical to semantic modeling.
Semantic Searching focuses on meaning and context rather than predetermined keywords. A semantic search engine can use artificial intelligence to identify query matches based on words and their context. Such a search engine can analyze by location, intent, word variations, synonyms, and concept matching. Requirements for semantic search involve figuring out what users want which means thinking like the users.
Web content optimized for semantics incorporates natural key words, rather than depending on rigid keyword insertion. Types of semantic keywords include: Core keywords that contain variations; thematic keywords for conceptually related terms; and stem keywords that anticipate what people might ask. Content can be further optimized through content relevancy and ‘shareworthiness’, and sharing content through social media integration.
Users of Business Intelligence (BI) and analytics tools often have semantic search requirements. The BI tools need to be flexible so that business users can find the information they need for analysis, reports and dashboards. Users of Big Data have a similar need to find common meaning in data from disparate formats.
