Entification

Entification is the process of giving an abstract "thing" (like a concept, data point, or person) a distinct, objective identity and structure, often by linking it to other entities to form a network, such as a knowledge graph. It transforms unstructured “strings” of information into structured “things” with defined relationships, making them actionable and usable in applications, thereby moving towards the goals of the Semantic Web. 

Key Aspects of Entification

• Moving from Strings to Things:  The core idea is to take raw, unstructured data (like text descriptions) and turn them into meaningful, identifiable entities. 
• Entity Identification and Reconciliation:  This involves identifying individual entities and ensuring they have a unique, consistent identifier, even if they appear in different forms or sources. 
• Relational Modeling:  Entification focuses on discovering and modeling the semantic relationships between these identified entities. 
• Operationalization:  Once entities are identified and related, they become operational in applications, such as knowledge graphs. 
• Knowledge Graph Construction:  Entification is a fundamental step in building knowledge graphs, where entities are nodes and their relationships are edges. 

Example

In the context of the Semantic Web, entification might involve:

• Taking a string of text like “Albert Einstein”. 
• Identifying it as a specific person and giving it a unique identifier. 
• Linking it to his known publications, his birth date, and his theories, creating a richer, interconnected “thing” rather than just a name. 

Application Areas

• Information Science:  Used to organize and structure information in databases and knowledge bases. 
• Archival Work:  Transforming archival metadata from simple “strings” to “things” to better link and model archival collections. 
• Marketing:  Brand entification refers to how consumers give human-like qualities to brands, leading to different user behaviors.